JP2006116565A - Welding system and welding robot control method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a consumable electrode type welding method by which the dead time of an arc generation sequence in a consumable electrode type welding method is eliminated without the occurrence of the buckling of a welding wire and spatters. <P>SOLUTION: A torch is moved in a direction of being separated from a base material 7 by a robot manipulator while a wire is being fed, and thus an actuator driving the robot manipulator can control the speed of the wire to the object to be welded by the operation in one direction of separating the torch from the base material, and vibration by the inversion of the torch speed is not generated. Further, by using a dedicated separation control system, the speed followability of the actuator moving the torch is increased without increasing overshoot in normal operation, and the acceleration-deceleration time of the manipulator can be reduced. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a control method in position control of a welding robot driven by a motor.

  In recent years, in the welding industry, efforts to further improve productivity have been made every day in order to ensure international competitiveness. In particular, demands for reducing so-called “choco stop”, which is a minor trouble that stops the production line, and shortening the tact time are increasing.

  There are various causes for the “choco stop”, but the biggest cause is a trouble caused by a defective arc start.

  For this reason, in the arc start in the conventional consumable electrode type welding method, when a start signal is input from the outside, the robot manipulator is moved to move the welding torch to the welding start position taught in advance, and then the wire With the robot manipulator stopped, the welding torch is moved approximately in the welding wire feeding direction to bring the tip of the wire closer to the workpiece, and it is determined that the tip of the wire has contacted the workpiece. Then, an initial current having a predetermined small current value is energized from the welding power source device, and the welding torch is moved in a direction opposite to the feeding direction of the welding wire to move the wire tip away from the workpiece. When the wire tip and the work piece are separated due to the backward movement, an arc is generated in which an initial current is applied. When the welding torch returns to the welding start position while maintaining it, the welding torch switches from the backward movement to the movement in the welding direction taught in advance, and simultaneously starts feeding the welding wire and energizes the steady welding current. It is known to shift from an initial arc generation state to a steady arc generation state by performing (see, for example, Patent Document 1).

  FIG. 4 is a schematic overall configuration diagram of a welding system using a robot that performs the above-described consumable electrode type welding method.

  In FIG. 4, reference numeral 101 denotes a welding wire that is a consumable electrode, which is fed from the wire spool 102 toward the welding torch 104 by the wire feed motor 103.

  Reference numeral 105 denotes a welding power source device. The welding power source device 105 supplies a predetermined welding current I and welding voltage V between the welding wire 101 and the base material 107 as a workpiece to be welded via the welding torch 104 and the welding tip 106. This is applied to generate an arc 108 and control the wire feed motor 103 to perform welding.

  A robot manipulator 109 holds the welding torch 104, positions the welding torch 104 (not shown), and moves the welding torch 104 along a welding line (not shown). .

  The robot manipulator 109 is controlled by the robot control device 110. The robot control device 110 performs two-way communication S with the welding power supply device 105, welding conditions such as a welding current I and a welding voltage V, A welding start or end command is transmitted.

  The consumable electrode type welding method in the welding system configured as described above will be described with reference to the time chart of FIG.

  FIG. 5 shows the respective states of the welding torch moving speed TV, the welding wire feed speed WF, the short-circuit determination signal A / S, the welding current I, and the welding voltage V in the vertical direction, and the horizontal axis represents time. The timing at which the welding start signal is transmitted from the robot controller 110 to the welding power source device 105 is TS0 ′, and TS1 ′ to TS5 ′ are the timings described later. .

  First, the robot control device 110 transmits a welding start signal to the welding power source device 105 and activates the robot manipulator 109 to accelerate the welding torch 104 toward the base material 107. The speed of the welding torch 104 becomes the initial torch speed. When TV0 is reached, the robot manipulator 109 stops accelerating and the welding torch 104 continues to descend at a constant speed.

  Further, when receiving the welding start signal from the robot control device 110, the welding power source device 105 applies a no-load voltage V 0 between the welding wire 101 and the base material 107.

  Eventually, at time TS1 ', when the welding wire 101 and the base material 107 come into contact with each other, a short-circuit determination signal A / S is output by a short-circuit determination unit (not shown) inside the welding power source device 105.

  The short-circuit determination signal A / S is transmitted to the robot controller 110 by the bidirectional communication S. The robot controller 110 immediately decelerates and stops the robot manipulator 109, and the operation of the robot manipulator 109, that is, the welding torch 104 at time TS2 ′. The speed becomes zero.

  Thereafter, the robot control device 110 immediately reverses the operation of the robot manipulator 109 and starts the operation in the direction in which the welding torch 104 moves away from the base material 107, and performs the lifting operation of the welding torch 104.

  Between this time TS1 ′ and TS3 ′ is an initial short-circuit period, and during this time, until the time TS2 ′ when the robot manipulator 109 decelerates and the speed becomes zero, the wire 101 is pressed against the base material 107. After the time TS2 ′, the operation of the robot manipulator 109 is reversed, so the pressing amount decreases, and the short circuit is released at the time TS3 ′.

  The timing of the time TS3 ′ at which the short circuit is released is greater in the area of the triangle cde that is the lifting amount than the area of the triangle abc that is the pressing amount of the welding wire 101 indicated by the TV line indicating the speed of the welding torch 104. Occurs at some point.

  The welding power source device 105 controls the welding current I to I1 'when an initial short circuit occurs at time TS1', increases the current to I2 'after a predetermined time, and waits for the short circuit to open.

  The reason for controlling the welding current to I1 ′ set at a relatively low level as the first stage of the initial short circuit period is that the welding wire 101 is melted by Joule heating of the tip of the welding wire 101 due to the initial short circuit, and the arc This is to prevent the welding wire 101 melted at the same time as being generated from being scattered and spattered.

  The reason for changing the current from I1 'to I2' is to give sufficient energy to generate an arc when the short circuit is released at time TS3 '.

  When an arc is generated at time TS3 ′, welding power supply apparatus 105 activates wire feed motor 103 to accelerate welding wire 101 toward base material 107, and the speed of welding wire 101 is the welding wire speed for main welding. After reaching the welding wire speed for main welding, the feeding of the welding wire 101 is continued at a constant speed.

Further, the welding power source device 105 controls the arc current I to the arc initial current I3 ′ for a certain period of time in conjunction with the start of the wire feed motor 103, and then controls it to the second initial current I4. (Not shown).
JP 2002-205169 A

  However, the conventional consumable electrode type welding method requires the reversing operation of the robot manipulator 109. That is, when it is determined that the tip of the welding wire 101 is in contact with the base material 107, the robot manipulator 109, which has moved forward, starts to decelerate, and after stopping once, the robot manipulator 109 reverses and accelerates in the reverse direction. In addition, acceleration / deceleration time is required.

  Since the robot manipulator 109 is generally driven by a motor via a speed reducer, when the speed is reversed by TS2 ′, the speed reducer play (meshing play), spring components, and friction discontinuity Vibration occurs due to the above.

  And this vibration may cause the welding wire 101 to be pressed against the base material 107 more than necessary, and there is a risk that the welding wire 101 will buckle.

  Further, even when the response of the robot manipulator 109 is poor and a delay occurs between the speed command of the welding torch 104 and the actual speed, the time from when the short circuit is detected at TS1 ′ until the speed is reduced and reversed at TS2 ′. There is a possibility that the welding wire 101 may be pressed more than necessary against the base material 107 as time elapses, and the welding wire 101 may be buckled.

  Here, it is the position control loop configured in the robot controller 110 that determines the responsiveness of the robot manipulator 109, and this configuration is shown in FIG.

  In FIG. 6, when the desired trajectory is operated by the welding torch 104, the trajectory of the motor position of each axis constituting the robot manipulator 109 is obtained by inverse kinematic calculation, and the speed component is set as the speed loop command TVC201 in the control loop. input.

  The reason why the input to the control loop is not the position component but the velocity component is that the position component requires a large number of significant digits including the entire robot operation range, and the data processing load becomes heavy.

  A torch position command TPC 203 obtained by integrating the torch speed command TVC 201 by the integration element 202 is input to the position loop 216.

  When teaching a trajectory to the robot manipulator 109, it is common to specify the positions of the start point and the end point, and the reason why the position loop 216 is necessary is that positioning cannot be performed accurately only with the speed loop 218.

  Therefore, in general feedback control (hereinafter referred to as FB control), a torch position command TPC 203 and a torch position feedback (hereinafter abbreviated as FB) signal output from the rotational position detection device 215 provided in the motor 214. The difference from the TPF 204 is multiplied by the position loop gain KPP 205 to generate a speed loop command TVCA 200 and input to the speed control loop 218.

速度制御ブロック２１８は、速度ループ指令ＴＶＣＡ２００とトーチ位置ＦＢ信号ＴＰＦ２０４を微分したトーチ速度ＦＢ信号ＴＶＦ２０８との差と、速度ループ比例ゲインＫＰＳ２０９と、積分要素２１０と、速度ループ積分ゲインＫＩＳ２１１とから、電流指令ＴＣＣ２１２を生成する。この電流指令ＴＣＣ２１２を受けた電流アンプ２１３が実際の電流をモータ２１４に流すことにより、マニピュレータ１０９を駆動する。

The speed control block 218 calculates the current from the difference between the speed loop command TVCA200 and the torch speed FB signal TVF208 obtained by differentiating the torch position FB signal TPF204, the speed loop proportional gain KPS209, the integral element 210, and the speed loop integral gain KIS211. A command TCC 212 is generated. The current amplifier 213 that has received the current command TCC 212 drives the manipulator 109 by causing an actual current to flow through the motor 214.

  In this state, generally, the speed loop command TVCA200 calculated by the position loop 216 is a signal whose phase is delayed from the torch speed command TVC201 which is a speed component of the motor position command TPC203. This is the main cause, and the torch speed FB signal TVF208 cannot sufficiently follow the torch speed command TVC201 and the phase is delayed. This is shown in FIG. 7 (a).

  FIG. 7A shows an example in which acceleration, constant speed, and deceleration are given to a general trapezoidal shape as the torch speed command TVC201. The torch speed FB signal TVF208 cannot sufficiently follow the torch speed command TVC201, the phase is delayed, and an error of a maximum of 35.3% occurs in the maximum speed ratio.

  Therefore, in order to improve the responsiveness of the robot manipulator 109, there is a system that adds a feedforward control (hereinafter referred to as FF control) shown in FIG.

  In FIG. 9, the FF control block 217 that multiplies the torch speed command TVC201 by the speed FF gain KFF219 and adds it to the speed loop command TVCA200 obtained by (Equation 1) in the position control loop 216 of the control block of FIG. Is added to the TVCB 206 as the speed loop command.

（数２）で計算されたＴＶＣＢ２０６を速度指令として速度制御ループ２１８に入力することで、後述するように、速度追従性が向上することが知られている。

It is known that by following the TVCB 206 calculated by (Equation 2) as a speed command to the speed control loop 218, the speed followability is improved as will be described later.

  The range of the speed FF gain KFF 219 is as shown in the following (Equation 3).

なお、図９において、ＫＦＦ＝０とすれば図６と等価である。

In FIG. 9, if KFF = 0, it is equivalent to FIG.

  Increasing the speed FF gain KFF 219, that is, approaching 1 improves the speed followability, but increases the overshoot. This is shown in FIGS. 7B and 7C.

  FIG. 7B shows a case where KFF = 0.5 and FIG. 7C shows a case where KFF = 1.0. As the torch speed command TVC 201, the same waveform as in FIG. 7A is given. Show.

  In the normal operation of the manipulator 109, the waveform shape of the torch speed command TVC201 is determined in advance from the load parameter of the manipulator 109, the maximum rotational speed of the motor 214, the maximum torque, and the like, and correction including the follow-up delay is performed in real time. It is common not to.

  8A to 8C are obtained by changing the vertical axis of FIGS. 7A to 7C from the torch speed to the torch position, and the end point position is 100%. Here again, as with the velocity, if the velocity FF gain KFF 219 is increased to approach 1, the position overshoot increases.

  And the increase of the position overshoot causes the following problems.

  After the welding is completed, the robot manipulator 109 moves to a retreat position that does not hinder the removal of the base material 107 that has been welded and the installation of the new base material 107, and approaches the base material 107 after the new base material 107 is installed. In general, the welding torch 104 is moved to a position at the time of TS0 ′ which is a welding start point. In this case, it is preferable that the end of the welding wire 101 stops without overshooting toward the base material 107 side. When overshooting occurs, the tip of the welding wire 101 comes into contact with the base material 7 at an unintended time, and the welding wire 101 may be bent or spatter may occur.

  That is, considering the movement from the retracted position to the welding start point TS0 ′, the speed FF gain related to the overshoot toward the base material 107 so that the tip of the welding wire 101 does not contact the base material 7 at an unintended time point. It can be seen that KFF 219 cannot be made so large.

  On the other hand, between TS0 ′ and TS2 ′ in the conventional arc start process shown in FIG. 5, as described later, the deceleration position indicated by TS1 ′ where the welding wire 101 contacts the base material 107 cannot be taught in advance. .

  That is, since the position of the short circuit is not constant due to the protruding amount of the welding wire 101 from the welding torch 104, the shape error of the base material 107, etc., the speed is reduced after the short circuit is detected by the short circuit determination signal A / S. Become.

  However, in the following, in order to simplify the description, it is assumed that there is no error in the protruding amount of the welding wire 101 and the shape of the base material 107.

  FIG. 10A shows a case where the speed FB signal TVF 208 has no follow-up delay between TS 0 ′ and TS 2 ′ in the conventional arc start process and completely follows the torch speed command TVC 201. FIG. 10B shows the position component on the vertical axis.

  10 (a) and 10 (b), at time TS1 ′ (= 0.4), the torch contacts the base material 107, a short circuit is detected, deceleration starts, and time TS2 ′ (= 0.0). It shall be stopped at 6).

  In the vertical axis of FIG. 10B, the movement amount up to time TS2 'is 100%, and the movement amount at the time of short-circuiting at time TS1' is 75%.

  However, actually, as shown in FIGS. 7A to 7C, the speed FB signal TVF208 has a follow-up delay with respect to the torch speed command TVC201. The time TS1 ′ at which 75% of 10 (b) is reached is delayed.

  FIG. 11A shows the relationship between the torch speed command TVC201 and the speed FB signal TVF208 when KFF = 0.

  In FIG. 11A, the torch speed command TVC201 is accelerated until time 0.2, and then becomes a constant speed. During this time, the speed FB signal TVF 208 accumulates errors due to the follow-up delay, and does not reach the short circuit position even at time 0.4 when the torch speed command TVC 201 reaches the short circuit position. Since no short circuit occurs, the torch speed command TVC201 does not decelerate and maintains a constant value. Thereafter, the time 0.47 (illustrated by TS1 ') is finally reached, and the speed FB signal TVF208 reaches the short-circuit position, so that the torch speed command TVC201 is decelerated to zero.

  However, after this, a follow-up delay occurs, and the torch descends until the speed FB signal TVF 208 decelerates to zero.

  FIG. 12A shows the vertical axis of FIG. 11A in terms of torch position, and the vertical axis shows the amount of movement up to time TS2 ′ in FIG. %.

  In FIG. 12A, the position FB signal TPF 204 has little overshoot with respect to the torch position command TPC 203, but the short-circuit time TS1 ′ is delayed due to the follow-up delay, so that the movement amount at the time TS2 ′ is , 17.6% increase with respect to the movement amount (100%) in FIG. That is, the amount of downward movement increases, and the wire 101 is excessively pushed into the base material 107, so that the wire 101 may be bent.

  FIG. 11C and FIG. 12C show waveforms when KFF = 1.0.

  In FIG. 11 (c), the speed overshoot is 13.3%, and the overshoot increases compared to when KFF = 0, but the follow-up delay is reduced and the maximum error is reduced to 13.3%. is doing.

  FIG. 12C shows the vertical axis of FIG. 11A in terms of torch position, and the vertical axis shows the amount of movement up to time TS2 ′ in FIG. 10B when there is no follow-up delay. 100%.

  In FIG. 12 (c), the position FB signal TPF204 has an overshoot with respect to the torch position command TPC203 which is increased as compared to when KFF = 0, but the short-circuit time TS1 ′ is almost delayed because the follow-up delay is small. As for the movement amount at time TS2 ′, the overshoot with respect to the movement amount (100%) in FIG. 10B is reduced to 3.1%.

  From the above results, between TS0 ′ and TS2 ′ in the conventional arc start process, since a short circuit is detected after TS1 ′ starts deceleration, even if the speed overshoot is large, the tracking delay is smaller, that is, KFF It can be seen that the amount of movement after the detection of the short circuit of TS1 ′ can be reduced and the risk of buckling of the welding wire 101 is reduced when setting = 1.0.

  However, if importance is placed on the speed followability required after the welding start point TS0 ′ and KFF = 1.0 is set, the amount of overshoot in the movement operation from the retracted position to the welding start point TS0 ′ increases. Therefore, the tip of the welding wire 101 comes into contact with the base material 107 at an unintended time, and there is a possibility that the welding wire 101 is buckled or spattered. Both speed improvement and suppression of overshoot can be achieved. It is difficult to adjust the FF gain KFF so as to satisfy.

  As described above, according to the conventional method, the welding wire 101 may be pressed against the base material 107 more than necessary due to the vibration at the time of direction reversal and the follow-up delay of the torch speed, so TS0 ′ to TS4 ′. The acceleration and deceleration of the torch in the meantime must be reduced. For this reason, there is a high possibility that the dead time due to the arc generation sequence becomes long.

  An object of the present invention is to provide a consumable electrode type welding method that can reduce the dead time of an arc generation sequence that a conventional consumable electrode type welding method has, without causing the welding wire to be bent or spattered. And

  In order to achieve the above object, a welding system of the present invention comprises a wire feeding means for feeding a welding wire to a welding torch, an actuator for holding the welding torch and moving the welding torch, and a position control system. A welding system having a control device for driving and controlling the actuator, and a welding power source for applying a welding output between the workpiece and a welding wire, the welding torch being welded by the actuator The control device controls the speed of the welding wire with respect to the workpiece to be separated from the workpiece, and the control device separates the actuator to move the actuator in the direction to separate the welding torch from the workpiece. A control system is provided.

  Further, the consumable electrode type welding method of the present invention includes a wire feeding means for feeding a welding wire to a welding torch, an actuator for holding the welding torch and moving the welding torch, and a position control system. The welding torch is covered by the actuator while feeding the welding wire using a welding system having a control device for controlling the actuator and a welding power supply device for applying a welding output between the workpiece and the welding wire. A consumable electrode type welding method that moves in a direction away from a workpiece and controls the speed of the welding wire with respect to the workpiece, wherein the controller separates the welding torch from the workpiece independently of the position control system. A dedicated separation control system for controlling the actuator in the direction is provided.

  According to this method, at the start of welding, the speed of the welding wire with respect to the work piece can be controlled by a one-way operation in which the actuator pulls the torch apart. By increasing the speed followability of the actuator that moves the welding torch using the control system, the response time and the acceleration / deceleration time can be shortened compared to the conventional one. In normal positioning, overshoot due to the separation control system is prevented by not performing control by the dedicated separation control system.

  As described above, the present invention has a conventional consumable electrode type welding method by moving the welding torch in the direction of separating from the workpiece by an actuator controlled using a dedicated separation control system at the start of welding. Reduce the dead time of the arc generation sequence that has been used to shorten the tact time, prevent the welding wire from buckling and spattering at the welding start, and effectively reduce the so-called "choco stop" Can do.

(Embodiment 1)
Hereinafter, an embodiment of the present invention will be described with reference to FIGS. 1 to 3, 7 and 8. FIG.

  First, the configuration of the welding system and the arc start process in the present embodiment will be described with reference to FIGS. 2 and 3, and then the robot control apparatus in the present embodiment will be described with reference to FIGS. 1, 7, and 8. The ten position control loops will be described.

  FIG. 2 is a configuration diagram showing an outline of the welding system in the present embodiment. Reference numeral 1 denotes a welding wire as a consumable electrode, which is fed from the wire spool 2 to the welding torch 4 by the wire feed motor 3. ing.

  Reference numeral 5 denotes a welding power source device. The welding power source device 5 has a predetermined welding current I and welding voltage between the welding wire 1 and the base material 7 to be welded via the welding torch 4 and the welding tip 6. V is applied to generate an arc 8 and the wire feed motor 3 is controlled to perform welding.

  Reference numeral 9 denotes a robot manipulator. The robot manipulator 9 holds the welding torch 4, positions the welding torch 4 (not shown), and moves the welding torch 4 along a welding line (not shown). It is.

  The robot manipulator 9 is controlled by a robot controller 10, which performs two-way communication S with the welding power source device 5 and performs various welding conditions such as a welding current I and a welding voltage V. Or, a welding start or end command is transmitted.

  FIG. 3 shows the situation of the moving speed TV of the welding torch 4, the feeding speed WF of the welding wire 1, the short-circuit determination signal A / S, the welding current I, and the welding voltage V in the vertical direction, and the horizontal axis This represents time, and as timing, the time when the welding start signal is transmitted from the robot control apparatus 10 to the welding power source apparatus 5 is denoted as TS0, and thereafter TS1 to TS5 represent respective timings. At time TS0, welding torch 4 is located at the welding start position. As will be described later, TS1 indicates the time at which the welding wire 1 and the base material 7 are in contact with each other, and TS2 indicates zero force applied to the direction in which the welding wire 1 is pressed against the base material 7, that is, the pulling speed of the welding torch 4 Shows the time when the feeding speed of the welding wire 1 is balanced, TS3 shows the time when the short-circuit between the welding wire 1 and the base material 7 is released, and TS4 is a welding torch where the height of the welding torch 4 becomes a predetermined height. 4 indicates the time when the pulling up of 4 is completed, and TS5 indicates the time when the wire speed WF becomes a predetermined speed.

  In FIG. 3, in this embodiment, when a welding start signal is transmitted from the robot controller 10 to the welding power source device 5 (TS0), the welding power source device 5 has no load voltage between the welding wire 1 and the base material 7. While applying V 0, the wire feed motor 3 is activated to accelerate the welding wire 1 toward the base material 7.

  When the feeding speed of the welding wire 1 reaches the initial wire speed W0, the acceleration of the wire feeding motor 3 is stopped and the wire feeding is continued at a constant speed.

  When the welding wire 1 and the base material 7 come into contact with each other at the time TS1, the short-circuit determination signal A / S is output by a short-circuit determination unit (not shown) in the welding power source device 5.

  This short-circuit determination signal A / S is transmitted to the robot control device 10 by the bidirectional communication S, and the robot control device 10 immediately starts the operation of the robot manipulator 9 in the direction in which the welding torch 4 is substantially separated from the base material 7, The lifting operation of the welding torch 4 is performed.

  The period between TS1 and TS3 is an initial short-circuit period. During this period, the welding wire 1 continues to be fed at the initial wire speed W0, and the robot manipulator 9 continues the pulling-up operation of the welding torch 4. As shown by the wavy line in the figure, the speed of the part is a speed obtained by combining the wire speed WF and the torch speed TV.

  Therefore, the tip of the welding wire 1 after TS1 is pressed against the base material 7 until TS2 where the composite speed indicated by the wavy line in the figure becomes zero, but the composite speed turns negative after TS2. The pressing amount decreases, and the short circuit is released at the time of TS3. TS3 is the time when the area of the triangle hji that is the lifting amount exceeds the area of the triangle fgh that is the pressing amount of the welding wire 1.

  It should be noted that welding power supply device 5 controls welding current I to I1 when an initial short circuit occurs at time TS1, increases the current to I2 after a predetermined time has elapsed, and waits for the short circuit to open.

  The reason for controlling the welding current I to I1 set relatively low as the first stage of the initial short-circuit period is that the welding wire 1 is melted by Joule heating at the tip of the welding wire 1 due to the initial short-circuit, and an arc is generated. This is to prevent the welding wire 1 melted simultaneously from being spattered.

  The reason for changing the current from I1 to I2 is to give sufficient energy to generate an arc when the short circuit is released at time TS3.

  In addition, in the subsequent pulling-up operation of the welding torch 4 after the short circuit is detected by TS1, a follow-up delay occurs in the speed TV of the welding torch 4, and the time during which the pulling-up speed TV of the welding tote 4 exceeds the speed WF of the welding wire 1, If it takes an excessive amount of time for the speed to turn negative, the welding wire 1 continues to be supplied during that time, so the pressing amount of the welding wire 1 becomes excessive, and the welding wire 1 may be bent.

  Further, similarly to the case where the deceleration position of TS1 ′ is not known in advance in the conventional example described in the problem to be solved by the invention, it is not possible to teach the short-circuit detection time of TS1 in advance. I can't do it.

  Therefore, in order to reduce the pressing amount of the welding wire 1, it is necessary to reduce the follow-up delay of the torch speed TV. Note that the welding method of the present embodiment is different from the conventional example in which the welding torch 104 moves to the base material 107 side from the position at time TS0 ′ described with reference to FIG. 5, and the welding torch from the position at time TS0. Since the operation direction 4 is only the direction away from the base material 7, there is no fear of buckling caused by vibration due to the change of the speed direction as shown in the conventional example.

  Moreover, since it is preferable that the composite speed of the welding torch 4 and the welding wire 1 is constant, the speed tracking error may be adjusted to be minimized.

  On the other hand, in the normal operation and positioning of the robot manipulator 9, the welding torch 4 that was in the retreat position for supplying the base material 7 at the time of TS 0 starts welding when the robot control device 10 drives the robot manipulator 9. Move to position.

  At this time, it is preferable to stop the tip of the welding wire 1 without overshooting toward the base material 7 side. This is because if the tip of the welding wire 1 overshoots, the tip of the welding wire 1 may come into contact with the base material 7 at an unintended time.

  However, since the trajectory accuracy, that is, the follow-up characteristic is required for the movement in the welding direction after TS4, it is preferable to increase the FF gain in a range where the overshoot is within an allowable value.

  Next, a position control loop in the present embodiment configured in the robot control apparatus 10 will be described.

  FIG. 1 is a block diagram showing a position control loop configured in the robot control apparatus 10. In FIG. 1, the same components as those in FIGS. 6 and 9 are denoted by the same reference numerals.

  The position control loop shown in FIG. 1 has a configuration in which a torch separation control block 224 for improving the response of only the torch pulling operation at the time of arc start is added to the position control loop of the conventional example described with reference to FIG. It has become.

  In FIG. 1, the speed commands are a torch pulling speed command TUVC222 relating to a welding torch pulling operation from time TS1 to TS4 shown in FIG. 3 at the start of welding, and a normal operation speed command TNVC223 for other normal operations.

  Then, the torch separation control block 224 performs FF (feed forward) control to improve the responsiveness of the operation of moving the welding torch 4 in the direction of separating from the base material 7 at the start of welding based on the torch pulling speed command TUVC222. Is to do. The torch pulling speed command TUVC 222 is a command that is output when the robot control device 10 performs control for moving the welding torch 4 in the direction of separating from the base material 7 at the start of welding, and is not output during normal operation. It is.

  The speed loop command TVCB206 output from the position control loop 216 to the speed loop 218 in the control loop of FIG. 1 will be described. The normal operation speed command TNVC 223 is added, and this addition is added to the speed loop command TCVA200 calculated in (Equation 1), whereby the speed loop command TVCB 206 is calculated. This TVCB is expressed by the following equation (Equation 4).

また、位置制御ループ２１６に出力される位置指令ＴＰＣ２０３について説明すると、位置指令ＴＰＣ２０３は、トーチ引き上げ速度指令ＴＵＶＣ２２２と通常動作速度指令ＴＮＶＣ２２３の和を積分したものである。この位置指令ＴＰＣを数式で表すと下記の（数５）となる。

The position command TPC 203 output to the position control loop 216 will be described. The position command TPC 203 is obtained by integrating the sum of the torch pulling speed command TUVC 222 and the normal operation speed command TNVC 223. This position command TPC is expressed by the following equation (5).

なお、ＫＦＦＡ＝ＫＦＦＢ＝０の時は、従来例として図６で示したＦＦ制御が無い制御ループと同じになる。

When KFFA = KFFB = 0, the control loop is the same as the control loop without FF control shown in FIG.

  Here, as described above, in the normal operation before TS0 and after TS4, it is preferable to increase the FF gain within a range where the overshoot is within an allowable value to improve followability.

  Therefore, when the speed FF gain KFFA220 for multiplying the normal operation speed command TNVC 223 is, for example, KFFA = 0.5, the following characteristics shown in FIGS. 7B and 8B are obtained, and the position overshoot is 1%. Although the following, the tracking characteristics are improved and the maximum position error is reduced to 9.2%.

  On the other hand, in the torch pulling operation from time TS1 to time TS4 shown in FIG. 3 at the start of welding, it is necessary to reduce the tracking delay of the torch speed TV and minimize the speed tracking error. Therefore, when the speed FF gain KFFB for multiplying the torch pulling speed command TUVC222 is, for example, KFFB = 1.0, the tracking characteristics shown in FIGS. 7C and 8C are obtained, and the speed tracking delay and the maximum error are obtained. Can be minimized.

  As described above, the position control loop of the robot control apparatus 10 is configured to be provided with the torch separation control block 224 shown in FIG. 1, and the torch pulling speed command TUVC 222 and the normal operation speed command TNVC 223 are separately input. The pulling speed command TUVC 222 is output only when the robot control device 10 performs control to move the welding torch 4 in the direction of pulling away from the base material 7 at the start of welding, and is not output during normal operation. Since the separation control block 224 can increase the speed followability of the movement of the welding torch 4, the dead time of the arc generation sequence can be reduced, and the welding wire 4 can be prevented from being bent or spattered.

  Further, since the torch pulling speed command TUVC 222 is not output during the normal operation, the normal operation can be performed without causing overshoot by the torch pulling control block 224 during the normal operation.

  That is, by configuring the position control loop shown in FIG. 1, it is possible to optimally adjust the speed tracking characteristics in the torch pulling operation from TS1 to TS4 at the start of welding and other normal operations.

  In the control method of the present embodiment, as described above, the operation direction of the welding torch 4 from the position at the time TS0 at the start of welding is only the direction away from the base material 7, and thus, as in the conventional example. There is no fear that the welding wire 1 will bend due to the vibration of the welding torch 4 due to the reverse of the speed direction of the welding tote 4.

  Further, when the arc start is performed while the welding torch 4 is operated in the welding direction, the welding direction operation and the torch pulling operation are performed at the same time. Since the torch pulling speed command TUVC222 relating to the welding torch pulling operation from time TS1 to TS4 shown in FIG. 3 and the normal operation speed command TNVC 223 in the other normal operations are separated, the speed tracking between the welding direction operation and the torch pulling operation is performed. Each characteristic can be adjusted optimally. In the method in which the FF gain is changed as a single speed command without separately setting the torch pulling speed command TUVC 222 and the normal operation speed command TNVC 223, it is difficult to appropriately change the FF gain. It is difficult to get started.

  The consumable electrode type welding method of the present invention reduces the tact time by reducing the dead time of the arc generation sequence that the conventional consumable electrode type welding method has, and at the same time, wire buckling and spattering at the welding start end. Since it is possible to prevent occurrence and effectively reduce the so-called “choco stop”, it is industrially useful as a consumable electrode type welding method used in, for example, production facilities and construction applications.

The block diagram which shows the position control loop in embodiment of this invention Schematic configuration diagram of a welding system used in an embodiment of the present invention Timing chart at the time of arc start in the embodiment of the present invention Schematic configuration diagram of welding system used in the prior art Timing chart at arc start in conventional technology Block diagram showing a position control loop in the prior art (A) A graph showing the relationship between the torch speed command and feedback when the feed forward gain is 0 in normal operation. (B) The torch speed command and feedback when the feed forward gain is 0.5 in normal operation. Graph showing relationship (c) Graph showing relationship between torch speed command and feedback when feed forward gain is 1 in normal operation (A) A graph showing the relationship between the torch position command and feedback when the feedforward gain is 0 in normal operation. (B) The torch position command and feedback when the feedforward gain is 0.5 in normal operation. Graph showing relationship (c) Graph showing relationship between torch position command and feedback when feedforward gain is 1 in normal operation Block diagram showing a position control loop with feedforward control in the prior art (A) Graph showing the relationship between the torch speed command and feedback when there is no follow-up delay in the torch pulling operation in terms of time and torch speed (b) Torch speed command and feedback when there is no follow-up delay in the torch pulling operation A graph showing the relationship in terms of time and torch position (A) Graph showing the relationship between the torch speed command and feedback in the torch pulling operation when the feedforward gain is 0, expressed as time and torch speed. (B) Torch pulling operation when the feedforward gain is 0.5. Graph showing the relationship between torch speed command and feedback in terms of time and torch speed (c) The relationship between the torch speed command and feedback in the torch pulling operation when the feed forward gain is 1, expressed by time and torch speed Graph (A) Graph showing the relationship between the speed command and feedback of the torch in the torch pulling operation when the feedforward gain is 0 in terms of time and torch position (b) Torch pulling operation when the feedforward gain is 0.5 (C) Graph showing the relationship between the torch speed command and feedback in terms of time and torch position in Fig. 3 (c) The relationship between the torch speed command and feedback in the torch pulling operation when the feed forward gain is 1 expressed in terms of time and torch position Graph

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Welding wire 3 Wire feed motor 4 Welding torch 5 Welding power supply device 7 Base material 8 Arc 9 Robot manipulator 10 Robot controller 200 Speed loop command TVCA
201 Torch speed command TVC
202 Integration element 203 Torch position command TPC
204 Differential element 205 Position loop gain KPP
206 Speed loop command TVCB
207 Differential element 208 Torch speed FB signal TVF
209 Speed loop proportional gain KPS
210 Integral element 211 Speed loop integral gain KIS
212 Current command TCC
213 Current amplifier 214 Motor 215 Rotational position detector 216 Position control loop 217 FF control block 218 Speed loop 219 Speed FF gain KFF
220 Speed FF gain KFFA
221 Speed FF gain KFFB
222 Torch pulling speed command TUVC
223 Normal operation speed command TNVC
224 Torch release control block

Claims (6)

A wire feeding means for feeding the welding wire to the welding torch, an actuator for holding the welding torch and moving the welding torch, a control device having a position control system for driving and controlling the actuator, A welding system including a welding power source for applying a welding output between a weld and a welding wire, wherein the actuator moves the welding torch in a direction away from the workpiece, and the speed of the welding wire with respect to the workpiece is increased. The control system includes a dedicated separation control system that moves the actuator in a direction to separate the welding torch from the work piece, separately from the position control system. The welding system according to claim 1, wherein the separation control system performs feedforward control. The welding system according to claim 1 or 2, wherein the separation control system is used only when the welding torch is moved in a direction in which the welding torch is separated from the workpiece. A wire feeding means for feeding the welding wire to the welding torch, an actuator for holding the welding torch and moving the welding torch, a control device having a position control system for driving and controlling the actuator, The welding torch is moved away from the work piece by the actuator using a welding system equipped with a welding power source device that applies a welding output between the work piece and the welding wire, and the speed of the welding wire relative to the work piece is increased. A consumable electrode type welding method for controlling, wherein the control device is provided with a dedicated separation control system for moving the actuator in a direction in which the welding torch is separated from the workpiece, separately from the position control system. Method. The consumable electrode welding method according to claim 4, wherein the separation control system performs feedforward control. 6. The consumable electrode type welding method according to claim 4 or 5, wherein the separation control system is used only when the welding torch is moved away from the workpiece.

Images