JP3166298B2 - Arc length control method for GMA welding - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a consumable electrode gas arc welding (hereinafter referred to as GMA welding) in which a consumable electrode is supplied and arc welding is performed using a shield gas containing an inert gas as a main component. In the method, the number of short circuits in which the tip of the consumable electrode short-circuits to the weld pool of the workpiece for a short time during welding is detected,
BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an arc length control method for GMA welding which obtains good welding quality by controlling the arc length to stabilize the arc.

[0002]

2. Description of the Related Art Generally, a GM for arc welding by supplying a consumable electrode using a shield gas mainly containing an inert gas.
In the A-arc welding method, a change in arc length L generated between a tip 1a of the consumable electrode 1 (hereinafter, referred to as a wire tip) and the surface of the workpiece 2 and an arc voltage value Va. Is
There are the relationships shown in FIGS. 43 (A) to (D). FIG.
As shown in FIG. 3A, when the arc 3 flies from the wire tip 1a to the shortest distance 2a on the surface of the workpiece 2, the shortest distance (hereinafter referred to as an apparent arc length) L1 The arc characteristic line indicating the relationship between the arc length La (horizontal axis) and the arc voltage value Va (vertical axis) shown in FIG. The operating point of the LV is located at the position shown in (A), and the arc voltage value is Va1. As shown in FIG. 43 (B), when the arc 3 flies to the position 2b where the cathode spot of the arc such as dirt or oxide on the surface is likely to occur, the apparent arc length becomes as shown in FIG. In spite of the same L1 as in (A), the actual arc length is La2, and the operating point of the arc characteristic line LV in FIG. (D) is at the position shown in (B). , And the arc voltage value Va2. This means that if the arc voltage value is Va2, as shown in FIG.
This is equivalent to the fact that the apparent arc length which is the shortest distance from the actual arc length La2 is L2.

Therefore, the apparent arc length (L1, L2)
Is detected as a change in the arc voltage value (Va1, Va2), the arc voltage value Va is not the actual arc voltage despite the fact that the apparent arc length is the same L1. Since it becomes Va1 or Va2 depending on the length La1 or La2, it is not possible to detect the arc voltage value and accurately control the arc length to a constant value. That is, when the actual arc length changes from the state of FIG. 43 (A) to the state of FIG. 43 (B) and the arc voltage value increases from Va1 to Va2, the apparent arc length L1 Despite the fact that has not changed
(When the arc voltage value becomes Va2 because the apparent arc length has increased to L2 as shown in FIG. 43 (C), an attempt is made to return the apparent arc length from L2 to the original L1. As in the case of the control, the apparent arc length is erroneously controlled in a direction shorter than L1.

Due to such erroneous control, in high-speed welding in which the apparent arc length is shortened and a good welding result is obtained recently, a short circuit frequently occurs, and the arc is short.
Instability causes welding defects and excessive spatter. Also, the apparent arc length L and the actual arc length
The phenomenon that the cut length La does not match is liable to occur in metals such as aluminum and magnesium where an oxide film is easily formed.
When the negative polarity of the output terminal of the welding power source is opposite to the polarity when it is connected to the workpiece, arc cathode spots are likely to be generated on the oxide film, and the cathode spot moves on a new oxide film. The actual arc length tends to be larger than the apparent arc length. Conventionally, there has been proposed a method of correcting the apparent arc length fluctuation so as to control the apparent arc length to a constant value as much as possible. It is not easy to perform erroneous control and to detect the apparent arc length itself because of the strong arc light.

On the other hand, when the apparent arc length becomes shorter, the number of short circuits between the wire tip 1a and the surface of the workpiece 2 increases, and conversely, when the apparent arc length becomes longer, a short circuit occurs. In a certain range, the apparent arc length L and the number of short circuits Q per unit time are in a proportional relationship. Therefore, conventionally, a method of controlling the arc length by detecting the number of short circuits per unit time and performing feedback to control the output voltage value of the welding power source (hereinafter referred to as a known technique).
There is. In this known technique, a predetermined number of short circuits Q per unit time are fed back, and the number of short circuits Q
Is equal to the target number of short circuits Qru per unit time,
The welding voltage set value or wire feed speed is controlled.
In this known technique, the target number of short circuits Qru per unit time is a long cycle of several to several tens of times due to the principle of detecting an appropriate arc length from the number of short circuits. Therefore, in the known technique, it is necessary to set a predetermined constant welding voltage value control cycle for securing the steady stability of the arc to about several seconds, and the transient response time Ttr to the arc length fluctuation is required. However, there was a major drawback that it was large.

Therefore, as a proposal for solving this problem,
JP-A-57-52537 and JP-A-57-152373
(Hereinafter referred to as “prior art”) has been proposed. This prior art, in addition to the well-known technique of controlling the number of short circuits to control the arc length within an appropriate range, furthermore, in order to shorten the transient response time to variations in the arc length,
A technique for changing the welding voltage correction amount or the length of the welding voltage value control period in accordance with the average welding current change rate. When the average welding current change rate is large, one welding output set value When the rate of change of the average welding current is small, the cycle is lengthened to obtain the steady stability of the arc.

[0007]

However, in the above-mentioned prior art, when the average welding current does not fluctuate or does not become large, the welding voltage correction amount or the length of the welding output value control period is increased or decreased. The effect of this prior art cannot be exerted because the time required for controlling to a predetermined arc length cannot be shortened because the control cannot be performed or these values are too small.

(Explanation of FIG. 44) A first example in which the effect of the prior art cannot be exhibited is a case where the prior art is applied to an aluminum MIG arc welding method. The reason will be explained. In FIG. 44, an aluminum alloy wire A5 having a diameter of 1.2 (mm) is used.
183 is shielded by argon gas, and the wire feeding speed W5 = 500 (cm / min) and W7 = 700 (cm / min)
FIG. 5 is a diagram showing a relationship between the average value Ia (A) of the welding current and the average value Va (V) of the welding voltage or the arc length La (mm) at the time of feeding. When the wire feeding speed is W5, the average value of the welding voltage is 1 when the average value of the welding current is set to 88.5 (A) and the apparent arc length L11 is 3 (mm).
At the operating point Q11 of 7.0 (V), the set value of the welding current is kept as it is, and the output voltage of the welding power source is changed to make the apparent arc length La = 9 (mm). Then, the operating point becomes Q12, and the average value of the welding voltage is 19.5.
(V), but the average value of the welding current is 91.5
(A), where the apparent arc length variation Lt1 = 6 (m
m), the variation ΔIa = 0 of the welding current. When the wire feed speed is W7, the average value of the welding current is set to 1
When the average value of the welding voltage when the apparent arc length L21 is set to 3 (mm) and the average value of the welding voltage is at the operating point Q21 of 19.3 (V), the set value of the welding current When the output voltage of the welding power source is changed to make the apparent arc length La = 9 (mm), the operating point Q22 is obtained.
Although the average value of the welding voltage changes up to 22 (V), the average value of the welding current is 132 (A), which is an apparent change in the arc length Lt2 = 6 (mm). The change amount of the current is 12 (A), and the change rate is about 10 (%).

As described above, in MIG arc welding of aluminum, even if the arc length changes greatly, the rate of change of the average welding current is small.
It is not possible to shorten the time for controlling the arc length to be appropriate.

(Explanation of FIG. 51) In MIG arc welding of aluminum, the time required for a welding voltage value roughly set by applying the prior art to reach a predetermined appropriate welding voltage value was measured. FIG. 51 shows a diameter of 1.6 (m) using the prior art.
m), the aluminum alloy wire A5183 was shielded with argon gas to convert the aluminum material A5083 to MIG.
Welding current value I (A), welding voltage value V (V) and number of short circuits Q per unit time (times / second) when arc welding is performed
It is a figure which shows the time progress t (second) (horizontal axis) of (vertical axis). In the figure, the set value of the welding voltage is roughly set and the
The voltage value immediately after the start is 20 (V), the welding current value is 200 (A), and the number of short circuits per unit time Q = 40 (times /
Second), the average value of the welding voltage is increased to about a predetermined appropriate value 22 (V) with respect to the average value 200 (A) of the welding current by about 7 to automatically increase according to the prior art.
(Seconds). The target number of short circuits Qru per unit time was set to 5 (times / second). The reason for the long time is that the amount of change in the average welding current is small as described with reference to FIG.

(Explanation of FIG. 52) FIG. 52 shows the measurement of the time required to reach a predetermined appropriate welding voltage value by applying the prior art to the MAG arc welding method of mild steel instead of aluminum of FIG. did. The measurement conditions in the figure are for a diameter of 1.2 (mm).
Is a condition for shielding the mild steel wire YGW15 of above with a mixed gas of 20% of carbon dioxide gas and 80% of argon gas to perform MAG arc welding of mild steel. In the figure, the welding voltage value is 28 (V) immediately after the setting value of the welding voltage is roughly set to be an arc start, and the welding current value is a predetermined appropriate value 32 (V) for 300 (A). ), It takes about 5 (seconds) to automatically increase by the prior art.
Although the welding current value changes by 10 (A) compared to the case of aluminum in FIG. 51 described above, the amount of change in the welding current is small and it takes time to reach an appropriate welding voltage value.

(Description of FIG. 53) Also, during welding, the arc length changes due to a change in the surface condition of the work to be welded, a change in the wire feed speed and a change in the welding current value, and a change in the heat input to the work to be welded. Due to the disturbance, the cathode spot of the arc moves irregularly to the position where the oxide film where the arc easily occurs is left.
Not only the apparent arc length but also the actual arc length fluctuates. When the cathode point moves irregularly due to such disturbance and the arc length fluctuates, it is necessary to control so that the apparent arc length corresponds to the appropriate number of short circuits Qru per unit time. There is a transient response speed Ttr (second) at that time
Need to be faster.

By the way, it is very difficult to actually measure whether the transient response speed Ttr is slow or fast with respect to an irregular disturbance. Therefore, the transient response speed Ttr was measured by replacing the irregular movement of the cathode spot due to disturbance under the following conditions. That is, it is known that the cathode spot of the arc is generated only within the range shielded by the shield gas. This means that the so-called cleaning phenomenon of removing the oxide film by moving the cathode point of the arc,
It means that it changes with the flow rate of the shield gas. Therefore, the actual measurement of the transient response speed is performed by forcibly switching the flow rate of the shield gas during welding to forcibly move the cathode point of the arc to change the actual arc length. Was.

FIG. 53 shows a welding current when an aluminum alloy wire A5183 having a diameter of 1.6 (mm) is shielded with argon gas and fed at a constant speed, and MIG arc welding is performed using the conventional technique. It is a figure which shows the time progress of the value I (A), the welding voltage value V (V), and the number of short circuits Qtu (times / second) per unit time (vertical axis). In the figure, the target number of short circuits Qru per unit time is 5 (times / second), the shield gas flow rate is 15 (liters / minute), and the welding voltage value V is appropriate for the welding current value I = 200 (V). If the flow rate of the shield gas is forcibly switched to 30 (liter / minute) during stable welding at a value of 21 (V), the shield gas shield range is increased and the cathode spot is increased. It is possible to move to a distant place, and the cleaning width is increased.
Since the cut length becomes large and the welding current value slightly decreases, the number of short circuits Qtu per unit time greatly increases to about 20 (times / second).

However, in this prior art, since the control cycle of the welding voltage value or the welding output value is changed according to the rate of change of the average welding current, the change in the average welding current as described above is small. In this case, the transient response time Ttr for returning the arc length to a value corresponding to the target number of short circuits Qru = 5 (times / second) per unit time requires about 5 (seconds).

As shown in the actual measurement examples of FIGS. 51 to 53 described above, in the conventional method of changing the welding voltage value or the welding output value control cycle by the change rate of the average welding current, the average welding current is not changed. When the rate of change is small, the problem that the transient response speed is slow remains.

[0017]

SUMMARY OF THE INVENTION An arc length control in GMA welding in which arc welding is performed by using a shield gas containing an inert gas as a main component to feed a consumable electrode and perform arc welding. The number of short circuits Q during the n-th welding output value control period.
n, the number of short circuits Qn is divided by the length Tn of the welding output value control period to calculate the number of short circuits Qtu per unit time, and the number of short circuits Qtu and the target number of short circuits Qru per unit time are calculated. The welding output correction amount ΔPn + 1 in the (n + 1) th welding output value control period corresponding to the difference Qtu−Qru is calculated, and from the function f (Qtu−Qru) of the difference between the number of short circuits Qtu and the target number of short circuits Qru, The length Tn + 1 of the next (n + 1) th welding output value control period is calculated, and the welding output correction amount ΔPn + 1 is added to the welding output set value Pn during the nth welding output value control cycle Tn. , Pn + 1 = Pn + ΔP
The welding output value during the (n + 1) th welding output value control period is controlled using n + 1 as the output set value, and the counting of the number of short circuits Qn + 1 during the (n + 1) th welding output value control period is started. This is a method of controlling the arc length of GMA welding which is repeated until the energization is completed.

According to a second aspect of the present invention, there is provided an arc length control method for GMA welding in which a consumable electrode is supplied and arc welding is performed by using a shield gas containing an inert gas as a main component. Moving average period T immediately before the end of the welding output value control period
m is divided by the moving average period Tm to calculate the moving average value Qmu = Qmn / Tm, and the difference Qmu between the moving average value Qmu and the target short circuit number Qru per unit time is calculated. Calculate the welding output correction amount ΔPn + 1 in the (n + 1) th welding output value control period corresponding to −Qru, and calculate the function f (Qmu−Q) of the difference between the average value Qmu and the target number of short circuits Qru.
The length Tn + 1 of the next (n + 1) th welding output value control period corresponding to ru) is calculated, and the welding output correction amount ΔPn + 1 is added to the welding output set value Pn in the nth control period. The welding output value in the (n + 1) th welding output value control cycle Tn + 1 is controlled using Pn + 1 = Pn + ΔPn + 1 as the welding output set value, and the number of short circuits during the (n + 1) th welding output value control period Tn + 1 This is a method of controlling the arc length of GMA welding, in which the counting of the welding current is started and thereafter repeated until the application of the welding current is completed.

According to a third aspect of the present invention, there is provided an arc length control method for GMA welding in which a consumable electrode is supplied and arc welding is performed using a shield gas containing an inert gas as a main component. Moving average period Tm immediately before the end of the welding output value control period
Is divided by the moving average period Tm to obtain a moving average value Qmu = Q during the moving average period Tm.
mn / Tm is calculated, the number of short circuits Qn during the n-th welding output value control period is counted, and the number of short circuits Qn is divided by the length Tn of the n-th control period to obtain the number of short circuits Qtu per unit time. = Qn / Tn, and the (n + 1) th welding output correction amount ΔPn + 1 corresponding to the difference Qtu−Qru between the number of short circuits Qtu and the target number of short circuits Qru per unit time is calculated. Function f of the difference from the target number of short circuits Qru
The length Tn + 1 of the next (n + 1) th welding output value control period corresponding to (Qmu-Qru) is calculated, and the welding output correction amount Δ
Pn + 1 is added to the welding output set value Pn in the n-th control period Tn to control the welding output value in the (n + 1) th welding output value control period using Pn + 1 = Pn + ΔPn + 1 as the welding output set value. At the same time, the MAG arc welding arc length control method starts counting the number of short circuits during the (n + 1) th welding output value control period Tn + 1 and thereafter repeats until the welding current is terminated.

According to a fourth aspect of the present invention, there is provided an arc length control method for GMA welding in which a consumable electrode is supplied and arc welding is performed using a shield gas containing an inert gas as a main component. The number of short circuits Qn during the welding output value control period is counted, and the number of short circuits Qtu is calculated by dividing the number of short circuits Qn by the length Tn of the welding output value control period to calculate the number of short circuits Qtu per unit time. Difference Qtu from target short circuit number Qru per hour
The welding output correction amount ΔPn + 1 in the (n + 1) th welding output value control period corresponding to −Qru is calculated, and the number of short circuits Q
Using the difference ΔQt = Qtu−Qru and the rate of change dQtu = Qtu−Qtu−1 between tu and the target number of short circuits Qru as input conditions, the length Tn + 1 of the next (n + 1) th welding output value control period is determined by fuzzy inference. Is determined and the welding output correction amounts ΔPn + 1 and n
Welding output set value Pn in the second welding output value control period
To control the welding output value in the (n + 1) th welding output value control period using Pn + 1 = Pn + ΔPn + 1 as the welding output set value, and to control the number of short circuits Qn during the (n + 1) th welding output value control period. This is an arc length control method for GMA welding in which the counting of +1 is started and thereafter repeated until the welding current is supplied.

According to a fifth aspect of the present invention, there is provided an arc length control method for GMA welding in which a consumable electrode is supplied and arc welding is performed by using a shield gas containing an inert gas as a main component. Moving average period T immediately before the end of the welding output value control period
The moving average value Qmu = Qmn / Tm is calculated by dividing the number of short circuits Qmn in m by the moving average period Tm, and the difference Qmu−Qru between the moving average value Qmu and the target number of short circuits Qru per unit time is calculated. The welding output correction amount ΔPn + 1 in the (n + 1) th welding output value control period corresponding to the above is calculated, and the difference ΔQm = Qmu−Qru between the average value Qmu and the target number of short circuits Qru and the rate of change dQmu = Qmu−Qmu are calculated. As the input condition, the length T of the next (n + 1) th welding output value control period by fuzzy inference
n + 1 is calculated, the welding output correction amount ΔPn + 1 is added to the welding output set value Pn in the n-th control period, and Pn + 1 = Pn + ΔPn + 1 is set as the welding output set value to be n + 1.
In addition to controlling the welding output value in the welding output value control period of the first time, the counting of the number of short circuits during the welding time control period Tn + 1 of the (n + 1) th time is started, and thereafter, the GMA welding operation is repeated until the current supply of the welding current is completed. -Length control method.

According to a sixth aspect of the present invention, there is provided an arc length control method for GMA welding in which a consumable electrode is supplied and arc welding is performed using a shield gas containing an inert gas as a main component. Moving average period T immediately before the end of the welding output value control period
m is divided by the moving average period Tm to obtain a moving average value Qmu = Q during the moving average period Tm.
mn / Tm is calculated, the number of short circuits Qn during the nth welding output value control period Tn is counted, and the number of short circuits Qn is divided by the length of the control period Tn to obtain the number of short circuits Qtu per unit time.
= Qn / Tn, and n + corresponding to the difference Qtu-Qru between the number of short circuits Qtu and the target number of short circuits Qru per unit time.
The first welding output correction amount ΔPn + 1 is calculated, and the difference ΔQm = Qru− between the moving average value Qmu and the target number of short circuits Qru is calculated.
The length of the next (n + 1) th welding output value control period Tn + 1 is determined by fuzzy inference using Qmu and the change rate dQmu = Qmu-Qmu-1 as input conditions, and the welding output correction amount ΔPn + 1 is determined.
And the welding output set value Pn in the n-th control period is added to control the welding output value in the (n + 1) th welding output value control period using Pn + 1 = Pn + ΔPn + 1 as the welding output set value. This is an arc length control method of GMA welding in which counting of the number of short circuits during the first welding output control period Tn + 1 is started and thereafter repeated until the end of the welding current.

According to a seventh aspect of the present invention, as the moving average value Qmu calculated in the second, third, fifth and sixth aspects, the number of short circuits Qn in each extraction unit cycle ΔT in the moving average cycle Tm is set.
Is divided by the number of extractions m in Tm (m = Tm / ΔT) Qmu = (Qn1 + Qn2 +... Qnm) / m where Qn1, Qn2. The number of short circuits during the unit .DELTA.T is used, and .DELTA.T and Tm are obtained by substituting predetermined constant values, thereby simplifying the calculation.

An eighth aspect of the present invention relates to an arc length control method for GMA welding in which a consumable electrode is fed using a shield gas containing an inert gas as a main component to perform arc welding.
n, the number of short circuits Qn during the n-th welding output value control period is counted, and the number of short circuits Qn is counted as the elapsed time N from the beginning of the n-th welding output value control period every time the short circuit extraction unit time ΔT elapses. Divide by ΔT (N is the number of times the short circuit extraction unit time ΔT has elapsed) to calculate the average number of short circuits Qtu per unit time up to that time, and calculate the average number of short circuits Qtu and the target number of short circuits Qru per unit time. The function f (Qtu
-Qru), the optimum length Tc of the welding output value control period is calculated, and the calculated value Tc is shorter than a predetermined length To, or the elapsed time N · ΔT is the length Tn of the welding output value control period. When the value becomes equal to, the n-th welding output value control period is terminated, and a welding output correction amount ΔPn + 1 corresponding to a difference (Qtu−Qru) between the average number of short circuits Qtu and the target number of short circuits Qru is calculated. The sum Pn + 1 = Pn + ΔPn + 1 of the welding output correction amount ΔPn + 1 and the welding output set value Pn during the n-th welding output value control period as an output set value;
This is an arc length control method for GMA welding in which the calculated value Tc is set as the length Tn + 1 of the welding output value control period, the (n + 1) th welding output value control period is started, and thereafter repeated until the end of the welding current.

According to a ninth aspect of the present invention, there is provided an arc length control method for GMA welding in which a consumable electrode is supplied and arc welding is performed using a shield gas containing an inert gas as a main component. The number of short circuits Qn during the n-th welding output value control period of Tn is counted, and every time the short circuit extraction unit time ΔT elapses, the number of short circuits Qmn in the immediately preceding moving average cycle Tm is calculated. From the calculated value Qmn, the moving average Q
mu = Qmn / Tm is calculated, and the optimum length Tc of the welding output value control period corresponding to the function f (Qmu-Qru) of the difference between the average value Qmu and the target number of short circuits Qru per unit time is calculated.
The calculated value Tc is shorter than a predetermined length To, or an elapsed time N · ΔT (N is the number of elapsed short-circuit extraction unit times ΔT) from the beginning of the n-th welding output value control period is set. Length Tn of the n-th welding output value control period
Is reached, the n-th welding output value control period is terminated, and a welding output correction amount ΔPn + 1 corresponding to the difference (Qmu−Qru) between the moving average value Qmu and the target number of short circuits Qru is calculated. The sum P of the welding output correction amount ΔPn + 1 and the welding output set value Pn during the n-th welding output value control period.
n + 1 = Pn + ΔPn + 1 is set as the welding output set value, and the calculated value Tc is set as n + 1 as the length Tn + 1 of the welding output value control period.
This is an arc length control method for GMA welding in which the first welding output value control period is started and thereafter repeated until the current supply of the welding current is completed.

According to a tenth aspect of the present invention, a consumable electrode is supplied by using a shield gas containing an inert gas as a main component and the arc gas is supplied.
In the arc length control method of GMA welding, the number of short circuits Qn during the n-th welding output value control period of the length Tn is performed.
Is counted, and every time the short circuit extraction unit time ΔT elapses, the number of short circuits Qmn in the immediately preceding moving average period Tm is calculated, and the moving average period Qmu = Qmn / Tm based on the moving average period Tm and the calculated value Qmn. Is calculated, and a function f (Qmu-Q) of the difference between the moving average value Qmu and the target number of short circuits Qru per unit time is calculated.
ru), the optimum length Tc of the welding output value control period is calculated, and the calculated value Tc becomes shorter than a predetermined length To or the elapsed time N from the beginning of the n-th control period.
When the ΔT (where N is the number of times of the short-circuit extraction unit time ΔT) reaches the length Tn of the n-th welding output value control period, the n-th welding output value control period ends, and the short-circuit occurs. The number of times Qn is divided by the elapsed time N · ΔT to calculate the average number of short circuits Qtu per unit time, and the welding output correction corresponding to the difference (Qtu−Qru) between the average number of short circuits Qtu and the target number of short circuits Qru. The amount ΔPn + 1 is calculated, and the sum Pn + 1 = Pn + ΔPn + 1 of the welding output correction amount ΔPn + 1 and the welding output set value Pn during the n-th control period is set as the welding output set value. This is an arc length control method for GMA welding in which the value Tc is set to the length Tn + 1 of the welding output value control period, the (n + 1) th welding output value control period is started, and thereafter repeated until the end of the welding current.

An eleventh aspect of the present invention relates to an arc length control method for GMA welding in which a consumable electrode is fed by using a shield gas containing an inert gas as a main component to perform arc welding. The number of short-circuits Qn during the n-th welding output value control cycle is counted, and the number of short-circuits Qn is counted as the elapsed time N.times. Divide by ΔT (N is the number of times the short circuit extraction unit time ΔT has elapsed) to calculate the average number of short circuits Qtu per unit time in the period up to that time.
Δt between Qru and the target number of short circuits Qru per unit time Qru = Qtu
The optimum length Tc of the next (n + 1) th welding output value control period is determined by fuzzy inference using −Qru and the rate of change dQtu = Qtu−Qtu−1 as input conditions. Or the elapsed time N · ΔT
Terminates the n-th welding output value control period when becomes equal to the length Tn of the n-th welding output value control period,
A welding output correction amount ΔPn + 1 corresponding to a difference (Qtu−Qru) between the average number of short circuits Qtu and the target number of short circuits Qru is calculated, and the welding output correction amount ΔPn + 1 and the n-th welding output value control are calculated. Sum Pn + with welding output set value Pn during period
1 = Pn + ΔPn + 1 is set as the welding output set value, the determined value Tc is set as the length Tn + 1 of the welding output value control period, and the (n + 1) th welding output value control period is started. This is an arc length control method for GMA welding.

According to a twelfth aspect of the present invention, a consumable electrode is supplied by using a shield gas containing an inert gas as a main component and an arc is supplied.
In the arc length control method of GMA welding, the number of short circuits Qn during the n-th welding output value control period of the length Tn is performed.
Is counted, and every time the short circuit extraction unit time ΔT elapses, the number of short circuits Qmn in the immediately preceding moving average period Tm is calculated, and the moving average period Qmu = Qmn / Tm based on the moving average period Tm and the calculated value Qmn. Is calculated, and the difference ΔQm between the moving average value Qmu and the target number of short circuits Qru per unit time is ΔQm = Qmu−Qru
The optimum length Tc of the welding output value control period is determined by fuzzy inference using the input condition and the rate of change dQmu = Qmu-Qmu-1, and the determined value Tc becomes shorter than a predetermined length To or n. Elapsed time N from the beginning of the second control period
When the ΔT (N is the number of times of the short-circuit extraction unit time ΔT) reaches the length Tn of the n-th welding output value control period, the n-th welding output value control period ends, and the movement is performed. The difference between the average value Qmu and the target number of short circuits Qru (Qmu−
Qru) to calculate the welding output correction amount ΔPn + 1, and sum Pn + 1 = Pn + of the welding output correction amount ΔPn + 1 and the welding output set value Pn in the n-th control period Tn.
ΔPn + 1 is set as the welding output set value, the determined value Tc is set as the length Tn + 1 of the welding output value control period, and the (n + 1) th welding output value control period is started. This is an arc length control method.

According to a thirteenth aspect of the present invention, a consumable electrode is supplied by using a shield gas containing an inert gas as a main component and the arc gas is supplied.
In the arc length control method of GMA welding, the number of short circuits Qn during the n-th welding output value control period of the length Tn is performed.
Is counted, and every time the short circuit extraction unit time ΔT elapses, the number of short circuits Qmn in the immediately preceding moving average period Tm is calculated, and the moving average period Qmu = Qmn / Tm based on the moving average period Tm and the calculated value Qmn. Is calculated, and the difference ΔQm between the moving average value Qmu and the target number of short circuits Qru per unit time is ΔQm = Qmu−Qru
The optimum length Tc of the welding output value control period is determined by fuzzy inference using the input condition and the rate of change dQmu = Qmu-Qmu-1, and the determined value Tc becomes shorter than a predetermined length To or n. Elapsed time N from the beginning of the second control period
When the ΔT (where N is the number of times the short-circuit extraction unit time ΔT has elapsed) reaches the length Tn of the n-th welding output value control period, the n-th welding output value control period ends,
The average number of short circuits Qtu per unit time is calculated by dividing the number of short circuits Qn by the elapsed time N · ΔT, and the welding output correction corresponding to the difference Qtu−Qru between the average number of short circuits Qtu and the target number of short circuits Qru. Calculate the amount ΔPn + 1 and determine the sum Pn + 1 = Pn + ΔPn + 1 of the welding output correction amount ΔPn + 1 and the welding output set value Pn in the n-th control period Tn as the welding output set value. This is an arc length control method for GMA welding in which the value Tc is set to the length Tn + 1 of the welding output value control period, the (n + 1) th welding output value control period is started, and thereafter the welding current is repeatedly applied until the end of the welding current.

The invention of claim 14 is the invention of claims 9, 10, 1
As the moving average value Qmu calculated in 2 and 13,
The sum of the number of short circuits Qn in each extraction unit period ΔT in the moving average period Tm is calculated as the number of extractions i in Tm (i = Tm / Δ
Qni divided by T) = (Qn1 + Qn2 +... Qni) / i, where Qn1, Qn2.
m is a calculation that is simplified by substituting a predetermined constant value.

[0031]

【Example】

(Embodiment 1) FIGS. 8 and 12 show a first embodiment of the control method shown in FIG. 1 corresponding to claim 1 of the present invention, and will be described with reference to FIGS. 7, 8 and 12. The control method of item 1 will be described.

(Explanation of FIG. 7) FIG. 7 is a block diagram when the arc length control method of the present invention is applied to a DC arc welding control device. In FIG. 1, a commercial power supply AC is input and a welding current is supplied from a welding power supply circuit PS between an electrode tip 4 of a consumable electrode 1 and a workpiece 2 to generate an arc 3. The consumable electrode 1 is fed by a wire feed roller WR rotated by a wire feed motor WM. The average welding current setting circuit IR outputs an average welding current setting signal Ir for setting an average value Ia of the welding current determined by the wire feeding speed of the wire feeding motor WM. The wire feed control circuit WC outputs a voltage Wc to the wire feed motor WM with the signal Ir as an input. Welding voltage detection circuit VD
Detects the instantaneous value of the welding voltage and determines whether or not there is a short circuit QD.
And a welding voltage detection signal Vd is output to the welding voltage comparison circuit CM1. The welding current detection circuit ID detects a welding current instantaneous value and outputs a welding current detection signal Id including an output current application start signal and an output current application end signal to an input / output signal conversion circuit I / O described later. The short circuit presence / absence determination circuit QD receives the welding voltage detection signal Vd as an input, determines a short circuit, and outputs a short circuit presence / absence determination signal Qd to the short circuit interruption circuit WK.

The short-circuit interrupt circuit WK receives the short-circuit presence / absence discrimination signal Qd from the central processing circuit CP.
U (not shown) formed in U
Is added to the count value of. The short circuit extraction unit clock timer TM is a timer that determines a short circuit extraction (sampling) cycle. For example, the short circuit extraction unit time ΔT = 100
An extraction unit time signal (clock signal) Ck is output every (ms). In the following embodiment of the present invention, the length Tn (second) of one welding output value control period is selected to be an integer value of 1 to 10 times the short circuit extraction unit time ΔT.
0.1 to 1.0 (second). This integer value indicates the number Nt of short-circuit extraction units during one welding output value control period described later.

The ROM is a read-only storage circuit, and is a read-only storage circuit for setting signals having predetermined set values, for example, a welding voltage setting signal Vr corresponding to the average welding current setting signal Ir, a target number of short circuits Qru per unit time, and the like. The initial value and the like of each constant are read. The RAM is a write / read storage circuit that writes the average welding current setting signal Ir, the number Nt of short circuit extraction units during the output value control period Tn, the count value Q of the number of short circuits, each set value, each operation value, and the like. And reading is performed.

The analog / digital conversion circuit A / D is
The average welding current setting signal Ir is converted into a digital value and output to the input / output signal conversion circuit I / O. This circuit I / O further transmits an output signal of the CPU to a digital / analog conversion circuit D / A. The central processing circuit CPU includes a short-circuit interrupt circuit WK, a short-circuit extraction unit clock timer T
M, an input / output signal conversion circuit I / O, a read-only storage circuit ROM, and a write / read storage circuit RAM. These circuits are shown in FIGS.
It has a function shown in a flowchart of a routine for controlling the welding output value control period and the welding voltage set value for each time.

The CM1 receives the welding voltage detection signal Vd and the welding voltage setting signal Vn in the (n-th) welding output value control period (length Tn) calculated by the central processing unit CPU and receives the difference between the welding voltage setting signal Vd and the welding voltage setting signal Vn. An output value control signal Ps is output to a welding power supply circuit PS including a welding output value control circuit to control the arc voltage to be substantially equal to a set value. A method of controlling the length of the welding output value control period and the welding voltage set value according to the first embodiment will be described with reference to FIGS.

(Description of FIG. 8) Block 31 (NC reset) Resets the count value Qn of the number of short circuits of the short circuit counter NC formed in the central processing unit CPU to zero.

Block 32A (Initialization of Nt corresponding to Tn) The length Tn of the welding output value control period is initialized. A short circuit extraction unit time ΔT which is predetermined as the length Tn of the control period
The number of times Nt (= Tn / ΔT) is used, and the initial value of Nt is 1. That is, the short-circuit extraction unit clock timer TM counts the number of short-circuits for a predetermined unit time, for example, 100 (ms). Output Ck. The welding output value control period is controlled to a single welding output value within this period. The length of the control period is an integral multiple of the short-circuit extraction unit time ΔT, for example, 1
To 10 are selected, the length of this control period Tn
Instead of the number of short-circuit extraction units Nt during this control period
(Times) can be used, and this replacement simplifies digital processing.

Block 33 (Ir) The average welding current setting signal Ir is read from the A / D into the RAM. Block 34 (Qru) A target short circuit number (for example, several to several tens) setting signal Qru per unit time stored in the ROM in advance is stored in the ROM.
From the CPU. Block 35 (Initialization of Vr corresponding to Ir) The initial value Vo of the welding voltage setting signal Vr stored in the ROM in advance is read, and the welding voltage comparison circuit CM is read from D / A.
Read to 1. Block 36 (Id) If the output current energization start signal included in the welding current detection signal Id is absent, the process waits. Block 37 (TM) When the welding operation is started and the output current energization start signal becomes available, the short circuit extraction unit clock timer TM starts counting.

Blocks 38 and 39 (Qd and NC) One unit of clock timer TM (short circuit extraction unit time ΔT,
If it is within 100 ms, for example, there is no interruption of the timer. If there is an interruption of the short-circuit presence / absence determination signal Qd, 1 is added to the short-circuit number counter NC in block 39, and the number of short circuits Qn is counted as Qn = Qn + 1. . After this counting or if there is no interruption of signal Qd, block 3
7, these loops are repeated until the short circuit extraction unit time ΔT is reached.

Blocks 40 and 41 (Id, Nt) When the loop of blocks 37 to 39 described above is repeated and the short circuit extraction unit time ΔT is reached, as long as the welding operation continues, it is included in the welding current detection signal Id. Since there is no output current energization end signal, 1 is subtracted from the number Nt of short-circuit extraction units during the control period Tn. Block 42 (Nt = 0) If the number Nt of short-circuit extraction units is not 0, the above-described loops 37 to 39 are repeated until the next short-circuit extraction unit time ΔT of the clock timer TM is reached, and the number of short circuits is added. to continue. When Nt = 0 is reached, the routine of block 50A (the flowchart of FIG. 12) is entered. When the routine of the block 43A (Q → 0) 50A (the flowchart of FIG. 12) is completed, the number of short circuits Q of the short circuit counter NC is reset, the process returns to the block 37, and the output current supply end signal of the block 40 is input. The loop from block 37 to block 43A and the loop from block 50A, that is, the loop from block 51 to block 57 in FIG. Block 49 (End) When the output current energization end signal is input in Block 40, the process ends in Block 49.

(Description of FIG. 12) In the following description,
From the number of short circuits Qn counted within the time Tn of the n-th welding output value control period, the correction amount ΔVn + 1 of the welding voltage set value in the next (n + 1) th welding output value control period is obtained, and the welding output value is calculated. A case in which control is performed after correction will be described.

Block 51 (storage of Qn) When the number Nt of short-circuit extraction units of the block 42 in the flowchart of FIG. 8 described above reaches 0, the number of short-circuits (the n-th control The number of short circuits Qn) during the period is stored in the RAM.

Block 52 (Calculation of Nt) The number of short circuits Qn during the nth control period is changed to the number of short circuit extraction units Nn (Nn =
Tn / ΔT) to obtain the number of short circuits Qtu per unit time.
= Qn / Tn = Qn / Nn. Block 53A (Calculation of ΔVn + 1) Based on the difference between the number of short circuits Qtu per unit time and the target number of short circuits Qru per unit time, the correction amount ΔVn + 1 = K1 (Qtu− Qru). Block 54A (Calculation of Tn + 1) From the function of the difference between the number of short circuits Qtu per unit time and the target number of short circuits Qru per unit time, the length of the next (n + 1) th welding output value control period Tn + 1 = Calculate f (Qtu-Qru). Block 55: The correction amount ΔVn + 1 of the welding voltage set value is added to the welding output set value Vn in the control period Tn, and the welding voltage set value Vn + 1 = V in the (n + 1) th welding output value control period.
Calculate n + ΔVn + 1. Block 56 The number of short circuit extraction units Nt + 1 = Tn + 1 / ΔT in the next control period calculated in block 54A is set to Nt.

Block 57: The welding voltage set value Vn + 1 is output from D / A. The welding voltage set value Vn + 1 in the (n + 1) th control period described in the block 57 of FIG. 12 is output to the welding voltage comparison circuit CM1 of the control device of FIG. 7, and as described in FIG. Control the voltage. When the operation of this block 57 is completed, after resetting the short circuit counter NC in block 43A of FIG.
Then, the loop of blocks 37 to 43A and block 50A, that is, the blocks 51 to 57 of FIG. 12 is repeated. When the output current supply end signal of the block 40 in FIG.

(Embodiment 2) FIGS. 9 and 13 show a second embodiment of the control method shown in FIG. 2 corresponding to claim 2 of the present invention. The control method of item 2 will be described. The control device to which the control method of claim 2 is applied is the device shown in FIG. 7 as in the first embodiment. In the control method according to claim 1 described with reference to FIGS. 8 and 12, the length Tn of one welding output value control period is determined by the number of short circuits Qtu per unit time and the target number of short circuits Qru per unit time.
And it changed with the difference. For example, in the first embodiment, since the control period changes in ten steps from 0.1 to 1.0 (second), when the control period length Tn is 0.1 (second) and short, the correction amount ΔV of the welding voltage set value is used. Therefore, the welding output value can be controlled immediately, so that the response speed is high. However, when the length Tn of the control period is as short as 0.1 (second), the length Tn of the control period immediately responds even though the number of data in one control period is small, The correction amount of the welding voltage also responds, and these may be repeated to cause an unstable phenomenon. Therefore, the control method shown in FIG. 2 corresponds to the following control by setting a fixed moving average period Tm with a predetermined sampling period and using the moving average value Qmu of the number of short circuits as a variable. The length Tn of the period is determined so that the number of data is prevented from decreasing even when the length Tn of the control period is shortened as described above, thereby preventing the occurrence of an unstable phenomenon.

(Description of FIG. 9) In FIG. 9, blocks having the same functions as those in FIG. 8 are denoted by the same reference numerals as those in FIG. 8, and description thereof will be omitted.

Block 44A The number Q of short circuits during the short circuit extraction period ΔT is counted and stored in Qnm of the FIFO memory in the RAM. At this time, F1F0
The memory is shifted by one stage, the contents of Qn1 are discarded, and Q
The content of n2 is moved to Qn1, and similarly, Qn2 to Qnm-1 are moved one stage before, respectively, and the latest m data Qn1 to Qn
nm is stored. Block 44B (Calculation of Qmu) Qn1 to Qnm are read from the FIFO memory and the average value Qmu per unit time of the total number of short circuits in the moving average period Tm is obtained.
= (Qn1 + Qn2 +... + Qnm) / m (where m = Tm / Δ
T) is calculated every time the short-circuit extraction unit time has elapsed. Block 43B (Reset of Counter) Every time the calculation of the average value Qmu of the block 44B is completed, the number of short circuits Q of the short circuit counter NC is reset.

(Description of FIG. 13) In the following description,
From the moving average value Qmu counted during the moving average period Tm immediately before the end of the nth control period, the next n
Calculate the length Tn + 1 of the + 1st welding output value control period and the correction amount ΔVn of the welding voltage set value in that period,
The case where the welding output set value is corrected and controlled based on this value will be described. 13, blocks having the same functions as in FIG. 12 are denoted by the same reference numerals as in FIG. 12, and description thereof will be omitted.

Block 53B (Calculation of ΔVn + 1) The (n + 1) th welding corresponding to the difference Qmu−Qru between the moving average value Qmu and the target short circuit number Qru per unit time within the moving average period Tm. The voltage set value correction amount ΔVn + 1 = K1 (Qmu−Qru) is calculated. Block 54B (Calculation of Tn + 1) From the function of the difference between the moving average value Qmu of the number of short circuits and the target number Qru of short circuits per unit time, the length of the next (n + 1) th welding output value control period Tn + 1 = f (Qmu-Qru) is calculated.

The welding voltage set value Vn + in the (n + 1) -th control period described in the block 57 of FIG.
1 controls the arc voltage in the order described with reference to FIG.
When the operation of this block 57 is completed, the loop of blocks 37 to 50B is repeated in the order described in FIG.
When the welding current ends, the process ends at block 49.

(Embodiment 3) FIG. 10 and FIG.
Is a third embodiment of the control method shown in FIG. FIG. 10 is different from FIG. 9 described in the second embodiment in that a process of accumulating and storing the total number of short circuits during one control period is added as 44C between 44A and 44B. Exactly the same.

The control method according to claim 3 will be described with reference to FIGS. The control device to which the control method of claim 3 is applied is the device shown in FIG. 7 as in the first embodiment. In the control method according to claim 2 described with reference to FIGS. 9 and 13, the length of the welding output value control period and the variable value of the correction amount ΔVr of the welding voltage set value are both the short-circuiting moving average value Qmu in the moving average cycle Tm. Met. Since the variable value that determines the length Tn of the control period is the moving average value Qmu of the number of short circuits in the moving average period Tm, a sufficient number of data can always be secured when the period Tm is appropriately selected. Stability is improved over operation. However, in the second embodiment, since the variable value of the correction amount ΔV of the welding voltage set value is also determined by the short circuit count moving average value Qmu,
The fluctuating number of short circuits is smoothed within the moving average period Tm, and the correction amount ΔV of the welding voltage set value becomes a small value. As a result, the stability is improved as compared with the first embodiment. , The response speed may be slow.

Therefore, in the control method shown in the corresponding diagram of claim 3 in FIG. 3, a predetermined fixed number of short-circuit extractions Nm, that is, a moving average period Tm is set in the same manner as in claim 2, By setting the moving average value Qmu of short-circuit times as a variable value, the length of the next control period is determined so that even if the length of the control period is shortened, the occurrence of an unstable phenomenon caused by repetition is prevented. As the variable value of the correction amount ΔV of the welding voltage set value, an average value Qtu per unit time of the total number of short circuits during one control period is employed as in the control method of the first embodiment of FIG. This prevents the response speed from slowing down.

(Description of FIG. 14) In FIG.
Similarly to the description of 3, the length of the next (n + 1) th welding output value control period is calculated from the short-circuiting number moving average value Qmu counted during the moving average period Tm immediately before the end of the nth welding output value control period. And the correction amount ΔVn + 1 of the welding voltage set value in the next control period from the average value Qtu of the number of short circuits per unit time during the n-th control period Tn.
Is calculated, and the welding output value is corrected and controlled based on this value. In FIG. 14, FIG.
Alternatively, a block having the same function as in FIG.
Or, if the same reference numerals as those in FIG.

In each of the above embodiments, both the welding output set value and the length of the welding output value control period are determined by a predetermined function using the difference from the target average number of short circuits as a variable. . According to the fourth to sixth aspects of the present invention, the welding output set value is determined by the difference from the target of the number of short circuits in the same manner as in the previous embodiment in order to make the control more reliable. The length is determined by fuzzy inference using the difference in the number of short circuits and the rate of change as input conditions. In the following, only the method of determining the length of the welding output value control period is different from each of the above-described embodiments, and therefore a description will be given with reference to a flowchart of this portion.

(Embodiment 4) FIG. 15 shows a routine of a part for obtaining the length of the welding output value control period and the welding output value set value in the embodiment of the control method shown in FIG. 9 is a flowchart, which corresponds to block 50A in the flowchart of FIG. 8 for explaining the operation of claim 1. Other flowcharts are the same as the flowchart in FIG. Therefore, the invention of claim 4 is the same as that of FIG.
The block 50A of the flowchart of FIG.
Therefore, the description of the same parts as those in FIG. 12 will be omitted, and only the specific parts in FIG. 15 will be described.

Block 54C (Tn + by fuzzy inference)
1) The difference ΔQt = Qtu−Qru between the number of short circuits Qtu per unit time and the target number of short circuits Qru per unit time, and the rate of change dQtu = Qt
Using u-Qtu-1 (where Qtu-1 is the previous time, that is, the number of short circuits per unit time in the (n-1) th control period Tn-1) as an input condition, the (n + 1) th control period is obtained by fuzzy inference. Length Tn + 1 is determined. An example of fuzzy inference will be described later.

The welding voltage set value V output for the (n + 1) th control period output in the block 57 of FIG.
n + 1 is output to the welding voltage comparison circuit CM1 of the control device of FIG. 7, and controls the arc voltage as described with reference to FIG. When the operation of this block 57 is completed, after resetting the number of short circuits Q of the number of short circuits counter NC in the above-mentioned block 43A of FIG. 8, the process returns to the block 37 of FIG. 8, and returns to the blocks 37 to 43A and the block 51 of FIG. To 57 are repeated. When the output current supply end signal of the block 40 in FIG.

(Embodiment 5) FIG. 16 mainly shows a flowchart of a part for determining the length of the welding output value control period in the embodiment of the control method shown in the corresponding diagram of claim 5 in FIG. This corresponds to block 50B in the flowchart of FIG. 9 for explaining the operation of the above-described claim 2.
Other flowcharts are the same as the flowchart in FIG. Therefore, the operation of the invention according to claim 5 corresponds to the flow chart of FIG. 9 in which the block 50B is replaced with FIG. 13 to FIG.

In FIG. 16, blocks 53B, 55 to 57 are blocks having the same functions as in FIG. 12, and therefore, are denoted by the same reference numerals as in FIG. 12, and description thereof will be omitted. Block 54D (Calculation of Tn + 1 by fuzzy inference) Difference ΔQm between moving average value Qmu of the number of short circuits and the target number of short circuits Qru immediately before the end of the n-th welding output value control period =
Qmu-Qru and the rate of change dQmu = Qmu-Qmu-1 (where Qmu
-1 determines the length Tn + 1 of the (n + 1) -th control period by fuzzy inference, using the moving average value immediately before the calculation of Qmu) as an input condition.

The welding voltage set value V in the (n + 1) -th control period output in the block 57 of FIG.
n + 1 controls the arc voltage in the order described in FIG. When the operation of the block 57 ends, the loop of the blocks 37 to 39 and the block 50B, that is, the loop of the blocks 53B to 57 of FIG. 16 is repeated in the order described with reference to FIG.

(Embodiment 6) FIG. 17 mainly shows a flowchart of a part for determining the welding output value control period in the embodiment of the control method shown in FIG. This flowchart corresponds to block 50C in the flowchart of FIG. 10 for explaining the operation of claim 3 described above. Other flowcharts are the same as the flowchart of FIG. Therefore, the operation of the invention of claim 6 corresponds to the operation in which the block 50C of the flowchart of FIG. 10 is replaced with that of FIG. 14 to FIG. In FIG. 17, the block 54B for determining the control period in FIG. 14 is replaced with the block 54D for determining by the fuzzy inference in FIG. The moving average Q of the number of short circuits counted within the time
From the mu, the length Tn + 1 of the next (n + 1) th welding output value control period is determined by fuzzy inference. A case is shown in which the correction amount ΔVn of the value is calculated, and the welding output value is corrected and controlled by this value. In FIG. 17, block 51
53 through 53A and 55 through 57 are the same as in FIG. 12, and the block 54D is the same as the block 54D in FIG.

In FIGS. 9 and 10, the average value Q per unit time of the total number of short circuits within the moving average period Tm is shown.
The block 44B that performs the operation of mu includes the extraction unit time ΔT
This calculation is performed every time the control period T elapses.
This may be performed at the end of n, that is, before Nt = 0 is determined in block 42 and the process branches to block 50B.
More specifically, FIG. 1 corresponding to block 50B of FIG.
The insertion position of the block 44B may be changed so as to be performed immediately before the block 53B of FIG. 3 or the block 54B of FIG. 14 corresponding to the block 50B of FIG.

Similarly, in FIGS. 16 and 17, the calculation of the moving average value Qmu may be performed at the latest before the block 53B in FIG. 16 or immediately before the block 54D in FIG. The resetting of the content Qn of the counter NC of the number of short circuits is performed by the block 4 in FIGS. 9 and 10, respectively.
The insertion position may be changed so that the insertion is performed between after 4A and after block 41.

(Example of fuzzy inference of control period length Tn + 1) Next, an example of fuzzy inference for determining the control period length Tn + 1 implemented in claims 4 to 6 will be described. Table 1 is an example of a control rule used in the present invention, in which the input condition is a difference ΔQ from the target value of the number of short circuits and a change rate dQ, and the conclusion is that the control period length Tn + 1 is obtained. . Here, the rate of change of the difference ΔQ / dt is as follows: ΔQ / dt = (ΔQn−ΔQn−1) / dt = {(Qn−Qru) − (Qn−1−Qru)} / dt = (Qn−Qn−1) / Dt Here, the number of short circuits per unit time by the current count is Qt
u, the number of short circuits per unit time of the previous count is Qtu-1, and when ΔQn, ΔQn-1 are represented by Qtu, Qtu-1, Qru, dt = 1, so that ΔQ / dt = Qtu−
Qtu-1 = dQ can be substituted and each claim 4
The values shown in Table 2 are used for ΔQ and dQ in the above-described items 6 to 6.

[0067]

[Table 1]

[0068]

[Table 2]

The reasons for defining the control rules as shown in Table 1 are as follows. However, in Table 1, Z is zero, N is Negat
ive, that is, ΔQ <0, dQ <0, P is Positive, that is, Δ
Q> 0 and dQ> 0. (1) When ΔQ = Z, dQ = Z When both the difference ΔQ and the rate of change dQ are zero (Z), the number of short circuits as set out and the state are calm.
The length Tn + 1 of the control period becomes longer (L). (2) When ΔQ = N, dQ = N ΔQ is N (Negative, ΔQtu <ΔQru), that is, the number of short circuits is smaller than the target value, and the rate of change dQ is also N (Negative ...
Since the number of short circuits is further decreasing, the change is abrupt and the length of the control period Tn + 1 is shortened (S). (3) ΔQ = P, dQ = P ΔQ is P (Positive, ΔQtu> ΔQru) and dQ is also P
, Tn + 1 is set to S because the number of short-circuits is large and conversely increasing. (4) When ΔQ = N, dQ = Z or ΔQ = P, dQ = Z In this case, the number of short-circuits is off the target, but the rate of change is zero, so it is calm in the off-target state. It indicates that. In this case, since ΔQ ≠ 0, the output set value Pn + 1 in the next control period is corrected to Pn + 1 = Pn + ΔPn + 1 corresponding to this ΔQ, so that the state naturally changes, and Also changes, so that Tn + 1 is set to S in order to detect this change quickly. (5) When ΔQ = Z, dQ = N or ΔQ = Z, dQ = P In this case, the difference from the target is zero, but the rate of change is not zero. It shows that it matched the target value. This indicates that the eyes are approached by the correction operation, and Tn + 1 is set to M (intermediate). (6) When ΔQ = P, dQ = N or ΔQ = N and dQ = P, the sign of the difference and the rate of change are opposite, so the number of short circuits is approaching the target, and the correction operation is in the convergence direction. Tn + 1 =
M is assumed.

FIG. 18 shows an example of a membership function of a triangle having three labels. In the figure,
(A) shows the difference ΔQ (ΔQt, ΔQm) between the number of short circuits and the target value.
), (B) show the rate of change dQ (dQtu, dQmu), (c)
Indicates the grade change of the control period length Tn + 1 obtained by applying the rule table 1 from these. Note that the membership function of the control period length Tn + 1 is set slightly to the left in order to improve the transient response, but this may be determined appropriately by experiments according to the response of other devices. Good.

Next, the procedure of fuzzy inference for determining the control period length Tn + 1 using Table 1 and FIG. 18 will be described with reference to FIG.
This will be described with reference to FIGS. Now, the difference ΔQ in the number of short circuits is +2 (slightly larger than the target value), and the rate of change dQ
Is +5 (increased 5 times). FIG.
Input condition ΔQ = + 2, dQ =
From FIG. 19 in which +5 is entered, ΔQ is Z = 0.6 and P =
0.1, and dQ overlaps at Z = 0.2 and P = 0.4. Therefore, from the rule table of Table 1 described above, ΔQ
(1) to (4) corresponding to each Z and P of dQ and dQ
Is true. Therefore, the fuzzy inference is shown in FIGS.
As shown in rule (1): (a) ΔQ (Z) = 0.6, (b) dQ
Since (Z) = 0.2, the logical product (MIN) of the two is taken, and (c) Tn + 1 (L) adopts 0.2 or less. Similarly, rule (2): (a) ΔQ (Z) = 0.6, (b) dQ
(P) = 0.4 to (c) Tn + 1 (M) = 0.4 or less, Rule (3): (a) ΔQ (P) = 0.1, (b) dQ
(Z) = 0.2 to (c) Tn + 1 (S) = 0.1 or less, Rule (4): (a) ΔQ (P) = 0.1, (b) dQ
From (P) = 0.4 to (c) Tn + 1 (S) = 0.1 or less, the conclusion is Tn + 1 (L) as shown by the shaded portion in each (c) of FIGS. = 0.2, Tn + 1 (M) =
0.4, Tn + 1 (S) = 0.1. Next, the logical sum (MAX) of the conclusion part of the inference result for each rule is shown in FIG.
become that way. The center of gravity of the hatched portion in FIG. 24 is calculated (diffusion processing) to obtain the control period length Tn + 1.
This length Tn + 1 is output as the length of the next control period.

Similarly, if the input condition is ΔQ = −3,
If dQ = + 5, ΔQ in FIG. 18 is N (0.6),
Since Z (0.1) and dQ overlap with Z (0.2) and P (0.4), rules (1),
(2), (5), and (6) are applied, and Tn +
1 (L) = 0.1, Tn + 1 (M) = 0.4, Tn + 1
(S) = 0.2, and the inference result is as shown in FIG. From the above result, the length Tn + 1 of the next control period is obtained, and during this period, the output set value Pn + 1 = Pn + ΔPn + 1
Controls the welding output.

(Description of FIG. 11) FIG. 11 shows the (n + 1) th welding output value control period Tn + 1 (vertical axis) calculated in block 54A of FIG. 12 in MIG welding of aluminum.
FIG. 8 is a function diagram of a control period showing a relationship between (second) and an absolute value (horizontal axis) of a value obtained by subtracting a target number of short circuits Qru per unit time from a number of short circuits Qtu per unit time. The length of the control period Tn + 1 in the present embodiment is as follows. When | Qtu−Qru | ≦ 10, Tn + 1 = 1.0−0.0
9 × | Qtu−Qru | | Qtu−Qru |> 10, Tn + 1 = 0.1

(Embodiment 7) FIG. 26 is a diagram corresponding to claim 8, and FIGS. 7, 32 and 35 show a seventh embodiment for implementing the control method according to claim 8, and FIG. 7, FIG. 32
The control method according to claim 8 will be described with reference to FIG. The control device to which the control method of claim 8 is applied is the device shown in FIG. 7 as in the first embodiment.

(Description of FIG. 32) Block 31 (NC reset) Resets the count value of the number of short circuits of the short circuit counter NC formed in the central processing unit CPU to zero.

Block 32A (initialization of Nt corresponding to Tn) The length Tc of the welding output value control period is initialized. A short circuit extraction unit time ΔT which is predetermined as the length of the control period Tc
The number of times Nt (= Tn / ΔT) is used, and the initial value of Nt is 1. That is, the short-circuit extraction unit clock timer TM counts the number of short-circuits for a predetermined unit time ΔT, for example, 100 (ms). ) Ck
Is output. The welding output value control period is controlled to a single welding output value within this period. The length of the control period is an integral multiple of the short-circuit extraction unit time ΔT, for example, 1
To 10 are selected, the length of this control period Tc
Instead of the number of short-circuit extraction units Nt during this control period
(Times) can be used, and this replacement simplifies digital processing.

Block 33 (Ir) The average welding current setting signal Ir is read from the A / D into the RAM. Block 34 (Qru) A target short circuit number (for example, several to several tens) setting signal Qru per unit time stored in the ROM in advance is stored in the ROM.
From the RAM. Block 35 (Initialization of Vr corresponding to Ir) Average welding current setting signal Ir stored in the ROM in advance
Is read out from the welding voltage setting signal Vr and output from the D / A to the welding voltage comparison circuit CM1. Block 36 (Id) If there is no output current conduction start signal included in the welding current detection signal Id, the process waits. Block 40 Checks the welding end command signal, and if there is an end signal, terminates all the control and moves to block 49;
Go to 7. Block 37 (TM) When the welding operation is started and the output current supply start signal is present, the short circuit extraction unit clock timer TM counts.

Blocks 38 and 39 (Qd and NC) One unit of clock timer TM (short circuit extraction unit time ΔT,
If it is within 100 ms, for example, there is no interruption of the timer. If there is an interruption of the short-circuit presence / absence determination signal Qd, 1 is added to the short-circuit number counter NC in block 39, and the number of short circuits Qn is counted as Qn = Qn + 1. . After this counting or if there is no interruption of the signal Qd, block 4
Returning to 0, these loops are repeated until the short circuit extraction unit time ΔT is reached.

Block 41 (Nt) When the loop of blocks 37 to 39 is repeated to reach the short circuit extraction unit time ΔT, 1 is subtracted from the number Nt of short circuit extraction units within the control period Tn, and FIG.
Enter the routine of the flowchart 50D. Block 43A (Q → 0) When the routine of the block 50D (the flowchart of FIG. 35 described later) ends, the short circuit count Qn of the short circuit counter NC is reset, and the process returns to the block 40 until the output current energization end signal is input. , The loop from block 37 to block 43A and the loop of the flowchart of FIG. Block 49 (End) When the output current energization end signal is input in Block 40, the process ends in Block 49.

(Description of FIG. 35) In the following description,
The optimum welding output value control period length Tc is calculated from the number of short circuits Qn counted within the elapsed time N · ΔT from the beginning of the nth welding output value control period at the end of the Nth short circuit extraction unit time ΔT. Then, the calculated length Tc is equal to a predetermined time To.
A case will be described in which the welding output value is corrected by obtaining the correction amount ΔVn + 1 of the welding voltage set value in the (n + 1) th welding output value control period when the period is shorter or when the period Tn has elapsed.

Block 51 (storage of Qn) When there is an interruption of the clock timer TM in the block 37 of the flowchart of FIG.
The number of short circuits (the number of short circuits Qn from the beginning of the n-th control period) counted by the short circuit counter NC is stored in the RAM.

Block 52 (Calculation of Qtu) The number of short circuits Qn is calculated as the length N up to the n-th control period.
.DELTA.T = Tn-Nt. Divide by .DELTA.T to obtain the number of short circuits Qtu = Qn / (Tn-Nt..DELTA.T) per unit time. Block 54E (Calculation of Tc) From the function of the difference between the number of short circuits Qtu per unit time Qtu and the target number of short circuits Qru per unit time, the optimal welding output value control period length Tc = f (Qtu-Qru) is determined. Calculate. Blocks 58 to 59 (Judgment of Tc and Judgment of Elapse of Tn) Either Tc calculated in block 54E is shorter than a predetermined time To, or the elapsed time N · ΔT is the length of the n-th welding output control period. When Tn is reached, the n-th welding output control period ends, and the routine proceeds to block 53A. T
When c> To and N · ΔT <Tn, block 3 in FIG.
Returning to 8, the counting of the number of short circuits is continued. Block 53A (Calculation of ΔVn + 1) Based on the difference between the number of short circuits Qtu per unit time and the target number of short circuits Qru per unit time, the correction amount ΔVn + 1 = K1 (Qtu− Qru). Block 55: The correction amount ΔVn + 1 of the welding voltage set value is added to the welding output set value Vn in the control period Tn, and the welding voltage set value Vn + 1 = V in the (n + 1) th welding output value control period.
Calculate n + ΔVn + 1. Block 56A The Tc calculated in block 54E is set to the length Tn + 1 of the welding output control period in the next control period (Tn + 1 = Tn).
c), the number of short circuit extraction units from now on Nt + 1 = Tc / ΔT
Is calculated and set to Nt.

Block 57: The welding voltage set value Vn + 1 is output from D / A. The welding voltage set value Vn + 1 in the (n + 1) -th control period described in the block 57 of FIG. 35 is output to the welding voltage comparison circuit CM1 of the control device of FIG. 7, and as described in FIG. Control the voltage. When the operation of the block 57 is completed, the short circuit counter NC is reset in the block 43A of FIG.
Returning to 0, the loop of the blocks 37 to 43A and the blocks 51 to 57 of FIG. 35 is repeated. When the output current supply end signal is input in block 40 of FIG. 32, the process ends in block 49.

(Embodiment 8) FIG. 27 is a diagram corresponding to claim 9, and FIGS. 33 and 36 show an eighth embodiment for implementing the control method of claim 9; The control method according to claim 9 will be described with reference to FIG. In addition,
The control apparatus to which the control method according to claim 9 is applied is Embodiment 7
This is the device shown in FIG. In the control method according to claim 8 described with reference to FIGS. 32 and 35, the length Tc of one welding output value control period is determined by the difference between the number of short circuits Qtu per unit time and the target number of short circuits Qru per unit time. It changed. For example, in the seventh embodiment, 0.1 to 1.0
(Sec) in 10 steps, so the length of the control period T
When c is 0.1 (second), which is short, the welding output value can be immediately controlled by the correction amount ΔV of the welding voltage set value, so that the response speed is high. However, when the length of the control period Tc is as short as 0.1 (second), the length of the control period Tc immediately responds even though the number of data in one control period is small, The correction amount of the welding voltage also responds, and these may be repeated to cause an unstable phenomenon. Therefore, the control method shown in FIG. 27 corresponds to the following control by setting a predetermined moving average period Tm having a predetermined sampling period and using the moving average value Qmu of the number of short circuits as a variable. The length Tc of the period is determined, and even when the length Tc of the control period is shortened as described above, a decrease in the number of data is prevented to prevent the occurrence of an unstable phenomenon.

(Description of FIG. 33) In FIG. 33, blocks having the same functions as in FIG. 32 are denoted by the same reference numerals as in FIG.

Block 44A The number of short circuits Q during one extraction unit time ΔT is counted, and R
It is stored in Qni of the FIFO memory in the AM. Then F
The 1F0 memory is shifted by one stage, the contents of Qn1 are discarded, the contents of Qn2 are moved to Qn1, and similarly, Qn2 to Qni
-1 are moved to the previous stage, respectively, and the latest i data Q
n1 to Qni are stored. Block 44B (Calculation of Qmu) Qn1 to Qni are read out from the FIFO memory and the average value Qmu per unit time of the total number of short circuits in the moving average period Tm is obtained.
= (Qn1 + Qn2 +... + Qni) / i (where i = Tm / Δ
T) is calculated every time the short-circuit extraction unit time has elapsed. Block 43B (Q → 0) Each time the calculation of the average value Qmu of the block 44B is completed, the number of short circuits Qn of the short circuit counter NC is reset.

(Description of FIG. 36) In the following description,
From the moving average value Qmu of the number of short circuits counted during the immediately preceding moving average period Tm at the end of the N-th short circuit extraction unit time, the optimal welding output value control period length Tc is calculated and calculated. When Tc is shorter than the predetermined time To or when the period Tn has elapsed, the correction amount ΔVn of the welding voltage set value in the (n + 1) th welding output value control period is calculated from the difference between the moving average value Qmu and the target value Qru. The case where the welding output set value is corrected and controlled based on this value will be described. In FIG. 36, blocks having the same functions as in FIG. 35 are denoted by the same reference numerals as in FIG. 35, and description thereof will be omitted.

Block 54F (Calculation of Tc) From the function of the difference between the moving average value Qmu of the number of short circuits and the target number Qru of short circuits per unit time, the optimal welding output value control period length Tc = f (Qmu-Qru ) Is calculated. Block 53B (Calculation of ΔVn + 1) The (n + 1) th welding voltage corresponding to ΔQm = Qmu−Qru, which is the difference between the moving average value Qmu and the target number of short circuits Qru per unit time within the moving average period Tm. The set value correction amount ΔVn + 1 = K1 (Qmu−Qru) is calculated.

The welding voltage set value V in the (n + 1) -th control period output in the block 57 of FIG.
n + 1 controls the arc voltage as described with reference to FIG. When the operation of this block 57 is completed, the loop of blocks 37 to 39 of FIG. 33 and blocks 54F to 57 of FIG. 36 is repeated in the same order as described with reference to FIG. 32, and a welding current energization end command signal is input. Then, the process ends at block 49.

(Embodiment 9) FIG. 28 is a diagram corresponding to claim 10, and FIGS. 34 and 37 are ninth embodiments for implementing the control method of claim 10. FIG. 34 shows a process of accumulating and storing the total number of short circuits during one control period in FIG. 33 described in the eighth embodiment.
4B and a block 44C, and the rest is exactly the same as FIG.

The control method according to claim 10 will be described with reference to FIGS. 34 and 37. The control device to which the control method of claim 10 is applied is the device shown in FIG. 7 as in the seventh embodiment. In the control method according to the ninth aspect, the variable value that determines the length Tc of the control period is a moving average period Tm.
Is the moving average value Qmu of the number of short circuits in
When an is selected appropriately, a sufficient number of data can always be ensured, and the stability is improved as compared with the operation of the seventh embodiment.
However, in the eighth embodiment, since the variable value of the correction amount ΔV of the welding voltage set value is also determined by the moving average value Qmu of the number of short circuits, the fluctuating number of short circuits is smoothed within the moving average period Tm. And the correction amount ΔV of the welding voltage set value.
Becomes a small value, and although the stability is improved as compared with the seventh embodiment, the response speed may be slowed down.

Therefore, in the control method shown in FIG. 28, a predetermined number of short-circuit extractions Nm, that is, a moving average period Tm is set in the same manner as in the ninth embodiment. By making the moving average value Qmu of short-circuit count a variable value, the length of the control period is determined to prevent the occurrence of instability caused by repetition of the control period even if the length of the control period is shortened. By using the average value Qtu per unit time of the total number of short circuits during one control period as in the control method of the seventh embodiment in FIG. The response speed is prevented from becoming slow.

(Explanation of FIG. 37) In FIG. 37, FIG.
6, the optimum welding output value control period Tc is determined from the short circuit count moving average value Qmu counted during the immediately preceding moving average period Tm at the end of the Nth short circuit extraction unit time ΔT. When the calculated length Tc becomes shorter than the predetermined value To or when the period Tn elapses, the n-th control period Tn ends, and the average value Qtu of the number of short circuits per unit time during this period is calculated. Is calculated, the correction amount ΔVn + 1 of the welding voltage set value in the next control period is calculated from this value, and the welding output value is corrected by this value to perform control. In FIG. 37, blocks having the same functions as those in FIG. 35 or FIG. 36 are denoted by the same reference numerals as those in FIG. 35 or FIG. The block 56B is the same as that shown in FIG.
This is obtained by adding a routine for resetting the count value Q of the total counter to 0 to the block 56A of FIG. 5 or FIG.

Also in each of the seventh to ninth embodiments, the length of the welding output value control period with respect to the absolute value of the value obtained by subtracting the target number of short circuits Qru per unit time from the number of short circuits Qtu or Qmu per unit time. The relationship shown in FIG. 11 is used for the relationship with Tc.

According to the eleventh to thirteenth aspects of the present invention, as in the fourth to the sixth aspects of the present invention, the welding output set value is set in the same manner as in the previous embodiment in order to more reliably control. The length of the welding output value control period is determined by a difference from the target of the number of short circuits, and the length of the welding output value control period is determined by fuzzy inference using the difference of the number of short circuits and the rate of change as input conditions. In the following, only the method of determining the length of the welding output value control period is different from each of the above-described embodiments, and therefore a description will be given with reference to a flowchart of this portion.

(Embodiment 10) FIG. 29 is a diagram corresponding to claim 11, and FIG. 38 obtains the length of the welding output value control period and the welding output value set value in the embodiment of the control method shown in FIG. It is a flowchart of a part of the routine, claim 8
32 corresponding to block 50D in the flowchart of FIG. Other flowcharts are the same as the flowchart in FIG. Therefore,
Since the invention of claim 11 corresponds to the block 50D of the flowchart of FIG. 32 replaced with FIG. 35 to FIG. 38, the same parts as those of FIG. Only specific parts will be described.

Block 54G (Tc by fuzzy inference
Each time the short circuit extraction unit time ΔT elapses, the difference ΔQt = Qtu−Qru between the number of short circuits Qtu per unit time from the beginning and the target number of short circuits Qru per unit time, and the rate of change dQtu = Qtu
Using -Qtu-1 (where Qtu-1 is the number of short circuits per unit time immediately before Qtu is calculated) as an input condition, an optimal control period length Tc is determined by fuzzy inference.
An example of fuzzy inference will be described later.

The welding voltage set value V output for the (n + 1) th control period output in the block 57 in FIG.
n + 1 is output to the welding voltage comparison circuit CM1 of the control device of FIG. 7, and controls the arc voltage as described with reference to FIG. When the operation of this block 57 is completed, after the number of short circuits of the number of short circuits counter NC is reset in the above-described block 43A of FIG. 32, the process returns to the block 40 of FIG.
The loop of blocks 37 to 43A and FIG.
The loop from block 51 to block 57 in FIG. 8 is repeated. When the output current supply end signal is input in block 40 of FIG. 32, the process ends in block 49.

(Embodiment 11) FIG. 30 is a diagram corresponding to claim 12, and FIG. 39 is a flowchart mainly showing a part for determining the length of the welding output value control period in the embodiment of the control method shown in FIG. This corresponds to the block 50E in the flowchart of FIG. 33 for explaining the operation of claim 9 described above. Other flowcharts are the same as the flowchart in FIG. Therefore, the operation of the twelfth aspect of the present invention is equivalent to the operation of the flowchart of FIG. 33 in which the block 50E is replaced with FIG. 36 to FIG.

In FIG. 39, blocks having the same functions as in FIG. 36 are denoted by the same reference numerals as in FIG. 36, and description thereof will be omitted. Block 54H (Calculation of Tc by Fuzzy Inference) Every time the short-circuit extraction unit time ΔT elapses, the difference ΔQm between the moving average value Qmu of the number of short-circuits immediately before and the target number of short-circuits Qru = Qm = Qmu
-Qru and the rate of change dQmu = Qmu-Qmu-1 (where Qmu-1
Is used as an input condition, and the optimal control period Tc is determined by fuzzy inference.

The welding voltage set value V in the (n + 1) th control period output in the block 57 of FIG.
n + 1 controls the arc voltage in the order described in FIG. When the operation of the block 57 is completed, the process returns to the block 40 of FIG.
The loop of blocks 54H to 57 of ninth is repeated, and the process ends at block 49.

(Embodiment 12) FIG. 31 is a diagram corresponding to claim 13, and FIG. 40 mainly shows a flowchart of a portion for determining the welding output value control period in the embodiment of the control method shown in FIG. This corresponds to block 50F in the flowchart of FIG. 34 for explaining the operation of claim 10 described above. Other flowcharts are the same as the flowchart in FIG. Therefore, the operation according to the thirteenth aspect of the present invention is equivalent to the operation of the flowchart of FIG. 34 in which the block 50F is replaced with that of FIG. 37 to FIG.

In FIG. 40, as in FIG. 37, the short-circuit count moving average value Qmu counted during the moving average cycle Tm immediately before the end of the N-th short-circuit extraction unit time ΔT is calculated as
The optimum welding output value control period Tc is determined by fuzzy inference, and when the calculated length Tc is shorter than the predetermined time To or when the period Tn has elapsed, n
A case is shown in which the correction amount ΔVn of the welding voltage set value in the next control period is calculated from the number of short circuits Qtu per unit time during the first control period, and the welding output value is corrected by this value to perform control. In FIG. 40, block 4
2A, 51, 52, 53A, 55 to 59 are the same as in FIG. 35, and block 54H is a block 54H in FIG.
Block 56B is the same as block 56 in FIG.
Since it is the same as B, detailed description is omitted.

(Example of Fuzzy Inference of Control Period Length Tc) Next, an example of fuzzy inference for determining the optimum control period length Tc implemented in claims 11 to 13 will be described. The control rules of Table 1 used in the preceding claims 4 to 6 are used as the control rules used in the inventions of Claims 11 to 13, and the difference ΔQ and the rate of change dQ from the target value of the number of short circuits are also the values shown in Table 2. Shall be adopted. Also, the membership functions when using the control rules shown in Table 1 are those shown in FIG.

When Table 1 and FIG. 18 are used, the procedure of the fuzzy inference for determining the length of the control period Tc is as described in the fourth embodiment.
In the sixth to sixth aspects, inference is performed in the same manner as the procedure described with reference to FIGS.

(Embodiment 13) FIG. 41 is a block diagram when the arc length control method of the present invention is applied to a pulse MAG arc welding control device. In the control circuit shown in FIG. 7, the welding voltage detection signal Vd and the welding voltage setting signal Vr are compared by the welding voltage comparison circuit CM1, and the welding output value is controlled by the difference welding output (voltage) value control signal Ps. Since the welding power supply control circuit PS was controlled, the welding power supply circuit was a constant voltage characteristic welding control circuit for directly controlling the welding output voltage value V. On the other hand, in the control circuit of FIG. 41, as will be described later, while the pulse voltage value Vp and the base voltage value Vb are both kept constant, the average value of the pulse current is controlled by controlling the pulse frequency f. The arc length is controlled to be substantially constant by changing the wire melting rate. In FIG. 41, components having the same functions as those of the control circuit in FIG. 7 are denoted by the same reference numerals, and description thereof will be omitted. Hereinafter, added components will be described. The pulse current value setting circuit IP1 sets a pulse current value and outputs a pulse current value signal Ip1. Base current setting circuit IB
1 sets a base current and outputs a base current setting signal Ib1. The pulse width setting circuit TP1 sets a pulse width and outputs a pulse width setting signal Tp1. Welding voltage comparison circuit CM
1 outputs a welding voltage control signal Cm1 representing the difference between the welding voltage detection signal Vd and the welding voltage setting signal Vr, as in FIG.
The pulse frequency control circuit VF receives the welding voltage control signal Cm1, and when the arc length increases and the number of short circuits decreases, the signal Cm1 increases and the pulse frequency control signal Vf output from the circuit VF increases. Conversely, when the arc length becomes small and the number of short circuits increases, this signal Cm1
Becomes small, and the pulse frequency control signal Vf output from the circuit VF becomes small.

The pulse width frequency control circuit DF is a circuit comprising a monostable multi-circuit that receives the pulse frequency control signal Vf and the pulse width setting signal Tp1 as components, and sets the pulse width at the same frequency as the pulse frequency Vf. A pulse width frequency control signal Df having the pulse width of the signal Tp1 is output. The pulse base current switching circuit SW1 includes a peak value determined by the pulse current value setting signal Ip1 and a pulse width frequency control signal Df.
A pulse control signal Pf is output by switching a signal corresponding to the pulse welding current having a pulse width determined by the above and a signal corresponding to the base current setting signal Ib1 at a pulse frequency f determined by the pulse frequency control signal Vf. The welding current comparison circuit CM2 compares the instantaneous value of the welding current detection signal Id with the pulse control signal Pf, and determines the difference between the welding output value control signal Ps and the welding output value control signal Ps.
Is output to the welding power supply control circuit PS including the control of the welding output value.

In FIG. 41, welding voltage control signal Cm1
The average value of the pulse current is changed by controlling the pulse frequency f, and the arc length is controlled to a substantially constant value by changing the wire melting speed.
The average value of the pulse current is changed by changing the pulse width TP or the pulse current value IP by controlling the pulse width TP or the pulse current value IP, or simultaneously controlling any two or these three of them, thereby changing the wire melting rate. The arc length may be controlled to a substantially constant value. Also, instead of controlling the base current value Ib to change the wire melting speed, and further controlling the welding output value to change the wire melting speed, the wire feed speed is changed to make the arc length substantially constant. The value may be controlled.

(Embodiment 14) FIG. 42 shows an arc length control method according to the present invention applied to a control device for controlling an arc length by controlling a wire feeding speed by controlling a wire feeding speed control circuit WS. It is a block diagram at the time of applying. In the first embodiment of FIG. 7 and the thirteenth embodiment of FIG. 41, the wire feeding speed is fed at a predetermined substantially constant value, and the output voltage value of the welding power source or the average value of the output current is controlled. In contrast to the method of controlling the arc length to a substantially constant value by increasing or decreasing the melting speed, the control method of FIG. 42 controls the wire feeding speed by controlling the wire feeding speed without controlling only the wire melting speed. In this method, the length is controlled to a substantially constant value.

In FIG. 42, components having the same functions as those in FIG. 7 are denoted by the same reference numerals, and description thereof will be omitted. Hereinafter, modified components will be described. In FIG. 42, a wire feed speed control circuit WS is used instead of the wire feed control circuit WC of FIG. 7, and this control circuit WS
, A welding voltage setting signal Vr output from the block 55 of the first to third embodiments of FIGS. 12 to 14 is input, and a wire feed speed signal Ws is supplied from the control circuit WS to the wire feed motor WM. Thus, the wire feeding speed is controlled so that the arc length becomes a substantially constant value.

In FIG. 42, a welding current comparison circuit CM2 is used in place of the welding voltage comparison circuit CM1 shown in FIG. 7, and the average welding current setting signal Ir and the welding current detection signal Id are compared. The value control signal Ps is input to the welding power supply control circuit PS to control the welding output current value to be substantially constant.

(Relationship Between Claims and Embodiments) The correction amount ΔPn + 1 of the welding output set value and the welding output set value according to claims 1 to 6 and 8 to 13. Pn or Pn + 1 is as follows in each embodiment. When controlling the welding output voltage value of the welding power supply control circuit as in the first embodiment of FIG. 7, the correction amount ΔVn + 1 of the welding voltage set value and the welding voltage set value Vn or Vn + 1 are respectively obtained. The pulse frequency, pulse width, and pulse frequency of the pulse MAG arc welding control device as shown in FIG.
When the pulse current value or the base current value is controlled, the respective set values are used and the average welding current changes according to these set values.
And the welding current set value In or In + 1. further,
When controlling the wire feed speed control device as shown in FIG. 42 of the fourteenth embodiment, the correction amount ΔFn + 1 of the wire feed speed set value and the wire feed speed set value Fn or Fn + 1 are respectively set.
Becomes

[0113]

【The invention's effect】

(Explanation of FIG. 45) FIG. 45 is a view showing the results of actual measurement when the methods of the first to sixth inventions of the present invention are carried out under the same welding conditions as in FIG. 51 of the prior art. That is, the measurement conditions were as follows: welding current value I (A), welding voltage value V (V) when aluminum material A5083 was MIG arc-welded by shielding aluminum alloy wire A5183 having a diameter of 1.6 (mm) with argon gas. FIG. 6 is a diagram showing the time lapse t (second) (horizontal axis) of the number of short circuits Qtu (times / second) (vertical axis) per unit time. In the figure, the set value of the welding voltage is roughly set and the welding voltage value immediately after the arc start is set to 20.
(V), when the welding current value is 200 (A) and the number of short circuits Qtu = 40 (times / second) per unit time,
The transient response time Ttr before automatically increasing the average value of the welding voltage to a predetermined appropriate value 22 (V) with respect to the average value 200 (A) of the welding current was reduced to about 1 (second). In such MIG welding of aluminum,
Compared with the prior art transient response time Ttr = 7 (second) of FIG. 51, the transient response time of the control method of the first to sixth aspects of the present invention of FIG. 45 is extremely large, Ttr = 1 (second). Shortening was realized.

(Explanation of FIG. 46) FIG. 46 is a diagram showing the results of actual measurements when the methods of the first to sixth inventions of the present invention are carried out under the same welding conditions as in FIG. 52 of the prior art.
That is, the welding conditions are as follows: a welding current value I (A) when a mild steel material YGW having a diameter of 1.2 (mm) is shielded with a mixed gas of 20% carbon dioxide gas and 80% argon gas to perform MAG welding on mild steel. ), The lapse of time t of the welding voltage value V (V) and the number of short circuits Qtu (times / second) per unit time (vertical axis)
It is a figure which shows (second) (horizontal axis). In the figure, the welding voltage value is 28 (V), the welding current value is 300 (A), and the number of short circuits Qtu = 20 (times / Second), the average value of the welding voltage is converted to the average value of the welding current of 300.
The transient response time Ttr until automatically increasing to a predetermined appropriate value 32 (V) for (A) is reduced to about 1 (second). Thus, even in the MAG welding of mild steel, compared with the transient response time Ttr = 5 (second) of the prior art of FIG. 52, the control method of the first to sixth aspects of the present invention of FIG. The transient response time was extremely shortened as Ttr = 1 (second).

(Explanation of FIG. 47) FIG. 47 is a diagram showing actual measurement results when the methods of the first to sixth inventions of the present invention are carried out under the same welding conditions as those of FIG. 53 of the prior art.
That is, welding conditions are as follows: an aluminum alloy wire A5183 having a diameter of 1.6 (mm) is shielded with argon gas, fed at a constant speed, and subjected to MIG welding using the control method of the present invention. A), with the welding voltage value V (V) and the target number of short circuits Qru = 5 (times / second) per unit time,
When the shielding gas flow rate is 15 (liter / min), the welding voltage value V is 2 which is an appropriate value for the welding current value I = 200 (A).
During the stable welding at 1 (V), the shielding gas flow rate was set to 30
(Liter / minute), the number of short circuits Qtu per unit time is 2 for the reason explained in FIG.
Although it greatly increases to about 0 (times / second), the transient response time Ttr required to return the arc length to a value corresponding to the target number of short circuits Qru = 5 (times / second) per unit time is about 1 (second). ). As described above, the transient response time of the first to sixth control methods of the present invention shown in FIG. 47 is also compared with the transient response time Ttr = 4 (second) of the prior art shown in FIG. Was able to realize a large shortening of Ttr = 1 (second).

(Description of FIG. 48) FIG. 48 shows the eighth to thirteenth embodiments of the present invention under the same welding conditions as those in FIG.
FIG. 7 is a diagram showing an actual measurement result when the control method of the invention is implemented. That is, in the figure, the set value of the welding voltage is roughly set and the arc is started.
(V), when the welding current value is 200 (A) and the number of short circuits Qtu = 40 (times / second) per unit time, the average value of the welding voltage is an appropriate value for the average value 200 (A) of the welding current. The response time until it automatically increases to 22 (V) is shown. In this case, To = 0.2 seconds as a fixed time for judging the end of the control period. As can be seen from the figure, the transient response time Ttr = 7 of the prior art of FIG.
As compared with (seconds), the transient response time according to the method of the eighth to thirteenth aspects of the present invention was Ttr = 0.7 (second), which was a very large reduction.

(Description of FIG. 49) FIG. 49 shows the eighth through thirteenth embodiments of the present invention under the same welding conditions as those in FIG. 52 of the prior art.
FIG. 7 is a diagram showing an actual measurement result when the control method of the invention is implemented. That is, in the same figure, the set value of the welding voltage is roughly set and the arc is started.
(V), when the welding current value is 300 (A) and the number of short circuits Qtu = 20 (times / second) per unit time, the average value of the welding voltage is an appropriate value for the average value of the welding current 300 (A). This shows a state until the voltage reaches 32 (V). Also in this case, To = 0.2 seconds as in FIG. As can be seen from the figure, even in the MAG welding of mild steel, the transient response time of the eighth to thirteenth methods of the present invention is Ttr = 5 in comparison with the transient response time Ttr of the prior art of FIG. 0.
An extremely large reduction of 7 seconds was achieved.

(Description of FIG. 50) FIG. 50 shows the eighth through thirteenth embodiments of the present invention under the same welding conditions as those in FIG.
FIG. 7 is a diagram showing an actual measurement result when the control method of the invention is implemented. That is, the target number of short circuits Qru = 5 per unit time
(Times / second), a predetermined time To to determine the end of the control period To =
0.2 seconds, shield gas flow rate 15 (liter / min),
This is a change when the shielding gas flow rate is forcibly switched to 30 (liter / min) during stable welding at a welding voltage value V of 21 (V), which is an appropriate value for the welding current value I = 200 (A). 32, the number of short circuits Qtu per unit time increases to about 20 (times / second), but the arc length is increased to the target number of short circuits Qru = 5 per unit time.
It shows a state until the value returns to a value corresponding to (times / second). As can be seen from the figure, the transient response of the control method according to the eighth to thirteenth aspects of the present invention is also larger than that of the prior art transient response time Ttr = 4 (seconds) of FIG. The time was greatly reduced to Ttr = 0.7 (second).

(Other effects) In the control method of the present invention, the hand-shake phenomenon, which is an effect of the prior art, that is, the distance between the electrode tip 4 and the workpiece 2 is caused by the vertical movement of the welding torch during semi-automatic welding. The number of short-circuits greatly changes even when abruptly changes. Therefore, the number of short-circuits that have changed suddenly and the target number of short-circuits are compared to change the correction amount ΔP of the welding output set value and / or the welding output value control period Tc. This also has the effect of shortening the transient response time Ttr for controlling the arc length to an appropriate value as in the prior art.
According to the arc length control method of the present invention, as described above, the transient response time for controlling the arc length to an appropriate value immediately after the arc start is reduced from 7 (seconds) or 4 (seconds) of the prior art to 1 (one).
(Second) to 0.7 (Second), so that the protruding of the wire tip at the welding start position, scattering of the wire tip, burnback, generation of spatter, excessive penetration, insufficient penetration, etc. are improved. We were able to.

The arc length control method of the present invention is particularly effective for MIG arc welding of aluminum or magnesium or an alloy thereof, in which a hard oxide film is easily formed. That is, as described above, when performing MIG welding of aluminum, since the cathode point of the arc moves in search of an oxide film when the polarity is reversed, even if the actual arc length fluctuates frequently, the apparent arc does not change. If the number of short circuits within a unit time does not change without changing the length, the welding output value control period is long, so the welding output value control is controlled over a long period, so that a hunting phenomenon does not occur and a stable arc continues. I do. Conversely, when the apparent arc length fluctuates, the welding output control is controlled in a short period of time, so that the transient response is good, and therefore, fluctuations in the wire feed speed, fluctuations in the wire protrusion length, etc. As a result, the appearance of the weld bead is good and the penetration depth is substantially constant.

Further, the arc length control method of the present invention employs a unified adjustment system in which the welding voltage value is automatically controlled to a predetermined appropriate value only by setting the welding current value. Setting can be done easily in a short time. Others
In high-speed welding of thin sheets of steel, stainless steel, etc., which have been spreading rapidly recently, the arc length is shortened and the allowable value of the arc length is narrowed. If it is not good, short-circuits frequently occur and spatters are likely to occur.However, the arc length control method of the present invention has excellent transient response and excellent stability, so that variations in arc length are extremely small. Therefore, it is possible to prevent an increase in spatter due to a long short circuit time or an excessive increase in the number of short circuits.

[Brief description of the drawings]

FIG. 1 is a diagram corresponding to claim 1;

FIG. 2 is a diagram corresponding to claim 2;

FIG. 3 is a diagram corresponding to claim 3;

FIG. 4 is a diagram corresponding to claim 4;

FIG. 5 is a diagram corresponding to claim 5;

FIG. 6 is a diagram corresponding to claim 6;

FIG. 7 is a block diagram when the arc length control method of the present invention is applied to a DC arc welding control device.

FIG. 8 is a half of a flowchart of the control method according to claims 1 and 4;

FIG. 9 is 1/1 of the flow chart of claim 2 and claim 5;
2.

FIG. 10 is a flow chart according to claims 3 and 6;
/ 2.

FIG. 11 shows a difference (horizontal axis) between the number of short circuits Qtu per unit time during the welding output value control period and the target number of short circuits Qru per unit time in the arc length control method of MIG arc welding of aluminum. , The length of the next welding output value control period Tn + 1
Or a control period function diagram showing a relationship with Tc (vertical axis).

FIG. 12 is a flowchart of the control method according to claim 1;
2.

FIG. 13 is a flowchart of a control method according to claim 2;
2.

FIG. 14 is a flowchart of a control method according to claim 3;
2.

FIG. 15 is a flowchart of the control method according to claim 4;
2.

FIG. 16 is a flowchart of the control method according to claim 5;
2.

FIG. 17 is a flowchart of the control method according to claim 6;
2.

FIG. 18 and FIG.
FIG. 14 is a diagram showing an example of a membership function applied in fuzzy inference used in な い し 13.

FIG. 19 is a diagram for finding an overlap when the input condition ΔQ = + 2 dQ = + 5 in the membership function of FIG. 18;

FIG. 20 is a diagram for explaining a process of fuzzy inference in rule (1) of Table 1.

FIG. 21 is a view for explaining the process of fuzzy inference in rule (2) of Table 1.

FIG. 22 is a view for explaining the process of fuzzy inference in rule (3) of Table 1.

FIG. 23 is a view for explaining the process of fuzzy inference in rule (4) of Table 1.

FIG. 24 is a diagram showing a result obtained by calculating a logical sum (MAX) of the results of rules (1) to (4).

FIG. 25 is a diagram showing the result of fuzzy inference when the input condition ΔQ = −3 dQ = + 5.

FIG. 26 is a diagram corresponding to claim 8;

FIG. 27 is a diagram corresponding to claim 9;

FIG. 28 is a diagram corresponding to claim 10;

FIG. 29 is a diagram corresponding to claim 11;

FIG. 30 is a diagram corresponding to claim 12;

FIG. 31 is a diagram corresponding to claim 13;

FIG. 32 is a flowchart of a control method according to claims 8 and 11;
1/2 of the chart.

FIG. 33 is a half of the flowchart of claims 9 and 12;

FIG. 34 is a half of the flowchart of claims 10 and 13;

FIG. 35 is a flowchart of a control method according to claim 8;
2.

FIG. 36 is a flowchart of a control method according to claim 9;
2.

FIG. 37 is a second flowchart of the control method according to claim 10;
/ 2.

FIG. 38 is a flowchart 2 of the control method according to claim 11;
/ 2.

FIG. 39 is a flowchart 2 of the control method according to claim 12;
/ 2.

FIG. 40 is a second flowchart of the control method according to claim 13;
/ 2.

FIG. 41 is a block diagram when the arc length control method of the present invention is applied to a pulse MAG arc welding control device.

FIG. 42 is a block diagram when the arc length control method of the present invention is applied to a wire feeding speed control device.

FIG. 43 is an explanatory diagram of a change in arc length.

FIG. 44 is an external characteristic diagram of an arc of aluminum MIG welding.

FIG. 45: When the arc length control method according to the first to sixth aspects of the present invention is applied to MIG arc welding of aluminum, an appropriate welding is performed based on a welding voltage value immediately after the arc start. The figure which shows the change until it reaches a voltage value.

FIG. 46 shows a case where the arc length control method according to any one of the first to sixth aspects of the present invention is applied to MAG arc welding of mild steel, from the welding voltage value immediately after the arc start to an appropriate welding voltage. The figure which shows the change until it reaches a value.

FIG. 47 shows a case where the first to sixth arc length control methods of the present invention are applied to MIG arc welding of aluminum, from when the cleaning width changes and the arc length changes. The figure which shows the change until the number of short circuits returns to the target number of short circuits Qru.

FIG. 48 shows a case where the arc length control method according to the eighth to thirteenth aspects of the present invention is applied to MIG arc welding of aluminum, and a proper welding is performed based on a welding voltage value immediately after the arc start. The figure which shows the change until it reaches a voltage value.

FIG. 49: When the arc length control method according to the eighth to thirteenth aspects of the present invention is applied to MAG arc welding of mild steel, an appropriate welding voltage is obtained from a welding voltage value immediately after the arc start. The figure which shows the change until it reaches a value.

FIG. 50: When the arc length control method according to the eighth to thirteenth aspects of the present invention is applied to MIG arc welding of aluminum, the cleaning width changes and the arc length changes. The figure which shows the change from the time until the number of short circuits returns to the target number of short circuits Qru.

FIG. 51 shows a conventional MIG arc of aluminum.
The figure which shows the change from the welding voltage value immediately after the arc start to the proper welding voltage value in arc welding.

FIG. 52 is a view showing a change from a welding voltage value immediately after the arc start to a proper welding voltage value in MAG arc welding of mild steel using a conventional technique.

FIG. 53 shows a conventional MIG arc of aluminum.
The figure which shows the change from the time when the cleaning width changes and the arc length changes until the number of short circuits returns to the target number of short circuits Qru in the arc welding.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 Consumable electrode 1a Tip of consumable electrode (wire tip) 2 Workpiece 4 Electrode tip AC commercial power supply PS (including control of welding output value) Welding power supply circuit WM Wire feed motor WC Wire feed control circuit WS wire Feeding speed control circuit IR average welding current setting circuit VD welding voltage detection circuit ID welding current detection circuit QD short circuit presence / absence judgment circuit CM1 welding voltage comparison circuit CM2 welding current comparison circuit WK short circuit interruption circuit CPU central processing circuit TM short circuit extraction unit clock Timer NC Short-circuit count counter ROM Read-only storage circuit RAM Write-read storage circuit ΔT Short-circuit extraction unit time Tn Length of n-th welding output value control period Tn + 1 Length of n + 1-th welding output value control period Tm Moving average period Tc Optimum welding output value control period length Qtu Number of short circuits per unit time Q mu Moving average value of the number of short circuits Qru Target short circuit number per unit time ΔQt Difference from the target value of the number of short circuits per unit time (Δ
Qt = Qtu−Qru) ΔQm Difference from the target value of the number of short circuits within the moving average time (ΔQmu = Qmu−Qru) dQtu Rate of change of the number of short circuits per unit time (= change from the target value of the number of short circuits per unit time) Change rate of difference: dQtu = Q
tu−Qtu−1) dQmu Rate of change of the number of short circuits within the moving average time (= change rate of difference from the target value of number of short circuits within the moving average time: dQmu)
= Qmu-Qmu-1) m Number of short circuit extraction units in the moving average period Tm (m
= Tm / ΔT) i The number of short-circuit extraction units in the moving average period Tm (i
= Tm / ΔT) N Number of elapsed short-circuit extraction units since the beginning of the n-th control period Nt Number of short-circuit extraction units during one control period (Nt
= Tn / ΔT) Pn Welding output set value in n-th control period Pn + 1 Welding output set value in n + 1-th control period ΔPn + 1 Correction amount of welding output set value for n + 1-th control period Vr Welding voltage set value Vo Initial value of welding voltage setting value Vn Welding voltage setting value in nth control period Vn + 1 Welding voltage setting value in n + 1th control period ΔVn + 1 Correction of welding voltage set value in n + 1th control period Tn + 1 Quantity Ttr Transient response time

──────────────────────────────────────────────────続 き Continuation of the front page (56) References JP-A-61-235077 (JP, A) JP-A-52-7844 (JP, A) JP-A-62-234663 (JP, A) JP-A-57-2784 22875 (JP, A) (58) Fields investigated (Int. Cl. ⁷ , DB name) B23K 9/12 B23K 9/095 B23K 9/173 G05B 13/02

Claims (14)

(57) [Claims] 1. An arc length control method for GMA welding in which a consumable electrode is fed by using a shield gas containing an inert gas as a main component and arc welding is performed, the n-th welding output value control period. The number of short circuits Qn during Tn is counted, and the number of short circuits Qn is counted.
Is divided by the length Tn of the welding output value control period to calculate the number of short circuits Qtu per unit time, and the (n + 1) th time corresponding to the difference Qtu−Qru between the number of short circuits Qtu and the target number of short circuits Qru per unit time The welding output correction amount ΔPn + 1 in the welding output value control period is calculated, and the following n is obtained from the function f (Qtu−Qru) of the difference between the number of short circuits Qtu and the target number of short circuits Qru.
Calculate the length Tn + 1 of the + 1st welding output value control period,
The welding output correction amount ΔPn + 1 is added to the welding output set value Pn during the n-th welding output value control cycle Tn to obtain P.
The welding output value in the (n + 1) th welding output value control period Tn + 1 is controlled using n + 1 = Pn + ΔPn + 1 as an output set value, and the number of short circuits Qn + 1 in the (n + 1) th welding output value control period is counted. Is started and thereafter repeated until the end of welding current application. 2. An arc length control method for GMA welding in which a consumable electrode is supplied and arc welding is performed using a shield gas containing an inert gas as a main component, and an n-th welding output value control. The number of short circuits Q during the moving average period Tm immediately before the end of the period
mn is divided by the moving average period Tm to calculate the moving average value Qmu = Qmn / Tm of the number of short circuits, and the (n + 1) th time corresponding to the difference Qmu-Qru between the moving average value Qmu and the target number of short circuits Qru per unit time The welding output correction amount ΔPn + 1 in the welding output value control period is calculated, and the next n corresponding to the function f (Qmu−Qru) of the difference between the average value Qmu and the target number of short circuits Qru is calculated.
Calculate the length Tn + 1 of the + 1st welding output value control period,
The welding output correction amount ΔPn + 1 is added to the welding output set value Pn in the n-th control period to obtain Pn + 1 = Pn +
Using ΔPn + 1 as the welding output set value, the welding output value during the (n + 1) th welding output value control period is controlled, and
GM which starts counting the number of short circuits during the + 1st welding output value control period Tn + 1 and thereafter repeats until the welding current is supplied.
Arc length control method for A welding. 3. An arc length control method for GMA welding in which a consumable electrode is fed by using a shield gas containing an inert gas as a main component and arc welding is performed, and an n-th welding output value control is performed. The number of short circuits Qmn during the moving average period Tm immediately before the end of the period
Is divided by the moving average period Tm to obtain the moving average period T
Calculate the moving average value Qmu = Qmn / Tm of the number of short circuits in m.
The number of short circuits Qn during the n-th welding output value control period is counted, and the number of short circuits Qn is determined by the length T of the n-th control period.
divided by n, the number of short circuits per unit time Qtu = Qn / Tn
To calculate the (n + 1) th welding output correction amount ΔPn + 1 corresponding to the difference Qtu−Qru between the number of short circuits Qtu and the target number of short circuits Qru per unit time, and calculate the moving average value Qmu and the target number of short circuits. The length Tn + of the next (n + 1) th welding output value control period corresponding to the function f (Qmu-Qru) of the difference from Qru
1 is calculated, and the welding output correction amount ΔPn + 1 is added to the welding output set value Pn in the n-th control period Tn, and Pn + 1 = Pn + ΔPn + 1 is set as n +
In addition to controlling the welding output value during the first welding output value control period, counting the number of short circuits during the (n + 1) th welding output value control period is started, and thereafter the arc length of GMA welding is repeated until the end of the welding current application. Control method. 4. An arc length control method for GMA welding in which a consumable electrode is fed by using a shield gas containing an inert gas as a main component to perform arc welding, the nth welding output value control cycle. The number of short circuits Qn during Tn is counted, and the number of short circuits Qn is counted.
Is divided by the length Tn of the welding output value control period to calculate the number of short circuits Qtu per unit time, and the (n + 1) th time corresponding to the difference Qtu−Qru between the number of short circuits Qtu and the target number of short circuits Qru per unit time , The welding output correction amount ΔPn + 1 during the welding output value control period is calculated, and the difference ΔQt = Qtu−Qru between the number of short circuits Qtu and the target number of short circuits Qru and the rate of change dQtu are calculated.
= Qtu-Qtu-1 as input conditions, the length Tn + 1 of the next (n + 1) th welding output value control period is determined by fuzzy inference, and the welding output correction amount ΔPn + 1 and the nth welding output value control are determined. The welding output set value Pn in the period is added, and P + 1 + 1 = Pn + ΔPn + 1 is used as the welding output set value, and n + 1
In addition to controlling the welding output value during the welding output value control period of the first time, the counting of the number of short circuits Qn + 1 during the n + 1th welding output value control period is started, and thereafter, the GMA welding arc is repeated until the current supply of the welding current is completed. Length control method. 5. An arc length control method for GMA welding in which a consumable electrode is fed by using a shield gas containing an inert gas as a main component and arc welding is performed, and an n-th welding output value control is performed. The number of short circuits Q during the moving average period Tm immediately before the end of the period
mn is divided by the moving average period Tm to obtain a moving average of the number of short circuits.
The value Qmu = Qmn / Tm is calculated, and the welding output correction amount ΔPn + 1 in the (n + 1) th welding output value control period corresponding to the difference Qmu−Qru between the moving average value Qmu and the target number of short circuits Qru per unit time is calculated. The difference ΔQm = Qmu-Qru between the average value Qmu and the target number of short circuits Qru and the rate of change dQmu = Qmu
−Qmu−1 and the next n by fuzzy inference
Calculate the length Tn + 1 of the + 1st welding output value control period,
The welding output correction amount ΔPn + 1 is added to the welding output set value Pn in the n-th control period to obtain Pn + 1 = Pn +
Using ΔPn + 1 as the welding output set value, the welding output value during the (n + 1) th welding output value control period is controlled, and
GM which starts counting the number of short circuits during the + 1st welding output value control period Tn + 1 and thereafter repeats until the welding current is supplied.
Arc length control method for A welding. 6. An arc length control method for GMA welding in which a consumable electrode is fed by using a shield gas containing an inert gas as a main component and arc welding is performed, the n-th welding output value control. The number of short circuits Q during the moving average period Tm immediately before the end of the period
mn is divided by the moving average period Tm to calculate a moving average value Qmu = Qmn / Tm during the moving average period Tm, and count the number of short circuits Qn during the n-th welding output value control period; The number of short circuits Qn is divided by the length of the control period Tn to calculate the number of short circuits per unit time Qtu = Qn / Tn, and the number of short circuits Qtu and the target number of short circuits Q per unit time Qtu are calculated.
The welding output correction amount ΔPn + 1 for the (n + 1) th time corresponding to the difference Qtu−Qru with respect to ru is calculated, and the difference ΔQm = Qmu−Qru between the moving average value Qmu and the target number of short circuits Qru and the rate of change dQmu
= Qmu-Qmu-1 as input conditions to determine the length Tn + 1 of the next (n + 1) th welding output value control period by fuzzy inference, and to determine the welding output correction amount ΔPn + 1 and the nth control period. The welding output set value Pn is added and Pn + 1 =
The welding output value in the (n + 1) th welding output value control period is controlled using Pn + ΔPn + 1 as the welding output set value, and counting of the number of short circuits during the (n + 1) th welding output control period is started. GMA to repeat until
Arc length control method for welding. 7. The moving average Qmu is a moving average period Tmu.
m = (Qn1 + Qn2 +... Qnm) / m = Qmn / The sum Qmn of the number of short circuits Qn in each extraction unit period ΔT in m is divided by the number of extractions m in the period Tm (m = Tm / ΔT).
m (where Qn1, Qn2,..., Qnm are the number of short circuits between the first to mth extraction units ΔT in Tm, and ΔT and Tm are predetermined constant values).
7. The arc length control method for GMA welding according to any one of 3, 5, and 6. 8. An arc length control method for GMA welding, in which a consumable electrode is fed by using a shield gas containing an inert gas as a main component and arc welding is performed, the n-th welding having a length Tn. It counts the number of short circuits Qn from the beginning of the output value control period, before
Short-circuit extraction unit from the beginning of the nth welding output value control period
Each time the time ΔT elapses, the number of short circuits Qn is
Elapsed time N · Δ from the beginning of the second welding output value control period
T (N elapsed number of short extraction unit time [Delta] T) is divided by
The average number of short circuits Qtu per unit time is calculated, and the optimum length Tc of the welding output value control period is calculated from the function f (Qtu-Qru) of the difference between the average number of short circuits Qtu and the target number of short circuits Qru per unit time. When the calculated value Tc becomes shorter than a predetermined length To or when the elapsed time N · ΔT becomes equal to the length Tn of the welding output value control period, the n-th welding output value control period ends. Then, the difference (Qtu-Q) between the average number of short circuits Qtu at the end and the target number of short circuits Qru is calculated.
ru), and calculates the sum Pn + 1 = Pn of the welding output correction amount ΔPn + 1 and the welding output set value Pn during the n-th welding output value control period.
+ ΔPn + 1 is set as the output set value, and the calculated value Tc is set as the length Tn + 1 of the welding output value control period, and the n + 1th welding output value control period is started. -Length control method. 9. An arc length control method for GMA welding in which a consumable electrode is fed by using a shield gas containing an inert gas as a main component to perform arc welding, the nth time of a length Tn. Each time the short-circuit extraction unit time ΔT elapses from the beginning of the welding output value control period, the number of short-circuits Qmn in the moving average period Tm immediately before the elapse of each short-circuit extraction unit time ΔT is calculated and calculated as the moving average period Tm. A moving average value Qmu = Qmn / Tm is calculated from the value Qmn and a function f (Qmu-Qru) of a difference between the average value Qmu and the target number of short circuits Qru per unit time.
Is calculated, and the calculated value Tc is shorter than a predetermined length To, or the elapsed time N · from the beginning of the n-th welding output value control period. When ΔT (N is the number of times the short-circuit extraction unit time ΔT has elapsed) reaches the length Tn of the n-th welding output value control period, the n-th welding output value control period is terminated. the moving average value Qmu with the calculated welding output correction amount .DELTA.Pn + 1 corresponding to the difference between the target number of short circuits Qru (Qmu-Qru), the welding output correction amount ΔP at time
The sum Pn + 1 = Pn + ΔPn + 1 of n + 1 and the welding output set value Pn during the n-th welding output value control period is set as the welding output set value, and the calculated value Tc is the length of the welding output value control period. An arc length control method for GMA welding in which the (n + 1) th welding output value control period is started as Tn + 1 and thereafter repeated until the welding current is not supplied. 10. GMA for arc welding by supplying a consumable electrode using a shield gas containing an inert gas as a main component.
In the arc length control method for welding , every time the short circuit extraction unit time ΔT elapses from the beginning of the nth welding output value control period of the length Tn , the short circuit extraction unit time ΔT elapses. The number of short circuits Qmn in the previous moving average period Tm is calculated, and the moving average period Qm = Qmn / Tm is calculated from the moving average period Tm and the calculated value Qmn. A function f (Qmu−
Qru), the optimum length Tc of the welding output value control period is calculated, and the calculated value Tc becomes shorter than a predetermined length To or the elapsed time N · ΔT from the beginning of the n-th control period. (n is the number of short extraction unit time [Delta] T) to exit the welding output value control period of the n-th when it reaches the length Tn of the n-th welding output value control period is set, before
Short from the beginning to the end of the nth welding output value control period
The number of entanglements Qn is counted, and the number of short circuits Qn is counted as the n-th welding.
The elapsed time N from the beginning to the end of the output value control period
Divide by ΔT to calculate the average number of short circuits Qtu per unit time, calculate the welding output correction amount ΔPn + 1 corresponding to the difference (Qtu−Qru) between the average number of short circuits Qtu and the target number of short circuits Qru, The sum of the welding output correction amount ΔPn + 1 and the welding output set value Pn during the n-th control period Pn + 1 = P
n + ΔPn + 1 is set as the welding output set value, and the calculated value Tc is set as the length Tn + 1 of the welding output value control period, and the n + 1th welding output value control period is started, and thereafter, GMA welding is repeated until the welding current is stopped. Arc length control method. 11. An arc length control method for GMA welding in which a consumable electrode is fed by using a shield gas containing an inert gas as a main component and arc welding is performed, the n-th welding having a length Tn. Count the number of short circuits Qn from the beginning of the output value control period,
From the beginning of the n-th welding output value control period,
Each time the delay time ΔT elapses, the number of short circuits Qn is increased by the elapsed time N · from the beginning of the n-th welding output value control period.
Divide by ΔT (N is the number of elapsed times of short circuit extraction unit time ΔT)
Calculates the average number of short circuits Qtu per unit time Te, as the difference ΔQt = Qtu-Qru the change rate dQtu = Qtu-Qtu-1 inputs the condition of the average number of short circuits Qtu unit time per target number of short circuits Qru An optimal length Tc of the next (n + 1) th welding output value control period is determined by fuzzy inference, and the determined value Tc becomes shorter than a predetermined length To, or the elapsed time N · ΔT is equal to the nth welding time. When the length of the output value control period becomes equal to the length Tn, the n-th welding output value control period is terminated, and the difference (Qtu-Qru) between the average number of short circuits Qtu and the target number of short circuits Qru at the end of this period. Is calculated, and the sum Pn + 1 = Pn + ΔPn + 1 of the welding output correction amount ΔPn + 1 and the welding output set value Pn in the n-th welding output value control period is calculated. The welding output set value is used, and the determined value Tc The n + 1 th welding output value control period starts as the length Tn + 1 value control period, GMA welding A repeated until power distribution end of the subsequent welding current - click length control method. 12. GMA for arc welding by supplying a consumable electrode using a shield gas containing an inert gas as a main component.
In the arc length control method for welding , every time the short circuit extraction unit time ΔT elapses from the beginning of the nth welding output value control period of the length Tn , the short circuit extraction unit time ΔT elapses. The number of short circuits Qmn in the previous moving average period Tm is calculated, and the moving average period Qm = Qmn / Tm is calculated from the moving average period Tm and the calculated value Qmn. Difference from target short circuit number Qru ΔQm = Qmu−Q
The optimum length Tc of the welding output value control period is determined by fuzzy inference using ru and the change rate dQmu = Qmu-Qmu-1 as input conditions, and the determined value Tc becomes shorter than a predetermined length To or When the elapsed time N · ΔT (N is the number of short-circuit extraction unit times ΔT) from the beginning of the n-th control period reaches the set length Tn of the n-th welding output value control period, the n-th welding output value control period starts. Exit of the welding output value control period, this
Output correction amount Δ corresponding to the difference (Qmu−Qru) between the moving average value Qmu and the target number of short circuits Qru at the end of the process
Pn + 1 is calculated, and the sum Pn + 1 = Pn + ΔPn + 1 of the welding output correction amount ΔPn + 1 and the welding output set value Pn in the n-th control period Tn is determined as the welding output set value. An arc length control method for GMA welding in which the value Tc is set to the length Tn + 1 of the welding output value control period, and the (n + 1) th welding output value control period is started and thereafter repeated until the end of the welding current. 13. GMA for arc welding by supplying a consumable electrode using a shield gas containing an inert gas as a main component.
In the arc length control method for welding , every time the short circuit extraction unit time ΔT elapses from the beginning of the nth welding output value control period of the length Tn , the short circuit extraction unit time ΔT elapses. The number of short circuits Qmn in the previous moving average period Tm is calculated, and the moving average period Qm = Qmn / Tm is calculated from the moving average period Tm and the calculated value Qmn, and the moving average value Qmu and the per unit time are calculated. Difference from target short circuit number Qru ΔQm = Qmu−Q
The optimum length Tc of the welding output value control period is determined by fuzzy inference using ru and the change rate dQmu = Qmu-Qmu-1 as input conditions, and the determined value Tc becomes shorter than a predetermined length To or When the elapsed time N · ΔT (N is the number of times the short-circuit extraction unit time ΔT has elapsed) from the beginning of the n-th control period reaches the set length Tn of the n-th welding output value control period, n The welding output value control period of the nth time is ended, and the welding output value control period of the n-th welding operation is started and ended.
And the number of short circuits Qn is counted as the n-th time.
Divided by the elapsed time N · ΔT from the beginning to the end of the welding output value control period, the average number of short circuits Qtu per unit time
The average number of short circuits Qtu and the target number of short circuits Qtu
welding output correction amount ΔPn + 1 corresponding to difference Qtu−Qru from ru
And the sum Pn + 1 of the welding output correction amount ΔPn + 1 and the welding output set value Pn in the n-th control period Tn.
= Pn + ΔPn + 1 as the welding output set value, and the determined value T
An arc length control method for GMA welding in which c is defined as the length Tn + 1 of the welding output value control period, and the n + 1th welding output value control period is started and thereafter repeated until the end of the welding current. 14. The moving average value Qmu is obtained by dividing the total of the number of short circuits Qn in each extraction unit time ΔT in the moving average period Tm by the number of short circuit extraction unit times ΔT in the period Tm i (i =
Tm / ΔT) Qmu = (Qn1 + Qn2 +... Qni) / i = Qmn / i (where Qn1, Qn2. The number of short circuits and Δ
14. The GM according to claim 9, wherein T and Tm are replaced by predetermined constant values.
Arc length control method for A welding.