GTAW liquid bridge transition closed-loop control system based on visual sense and electric arc cooperative sensing

Abstract

The invention relates to the field of GTAW filling wire control, in particular to a GTAW liquid bridge transition closed-loop control system based on vision and electric arc cooperative sensing. The GTAW liquid bridge transition closed-loop control system based on vision and electric arc cooperative sensing is a GTAW liquid bridge transition closed-loop control system based on vision and electric arc cooperative sensing. Aiming at keeping a continuous liquid bridge transition mode for welding in a GTAW process, the invention provides a GTAW liquid bridge transition closed-loop control system based on visual sense and electric arc cooperative sensing. A GTAW robot is used for controlling the welding action; shooting a GTAW droplet transition process by using a droplet transition image acquisition system; an arc voltage signal between the welding wire end and the workpiece and an arc voltage signal between the tungsten electrode tip and the workpiece are collected through an arc signal collecting system; identifying a liquid bridge transition mode by using a GTAW liquid bridge transition mode identification system; a filling wire height control system is utilized to enable a liquid bridge transition mode to be a continuous liquid bridge transition mode.

Description

Technical Field

[0001] The invention relates to the field of GTAW filling wire control, and is a GTAW liquid bridge transition closed-loop control system based on vision and arc cooperative sensing. Background Art

[0002] GTAW (gas shielded welding) is widely used in aerospace, marine equipment, nuclear power construction and other fields. Especially in recent years, with the rapid development of science and technology, higher requirements have been put forward for component performance, and GTAW has also developed in a more intelligent direction. During the GTAW welding process, maintaining a continuous liquid bridge transition mode for welding can improve the quality of forming. During the GTAW welding process, it may not be possible to maintain a continuous liquid bridge transition mode for welding. In response to the above difficulties, the present invention discloses a GTAW liquid bridge transition closed-loop control system based on vision and arc collaborative sensing, which makes the liquid bridge transition mode in a continuous liquid bridge transition mode by changing the filler wire height during the welding process. Summary of the invention

[0003] The GTAW liquid bridge transition closed-loop control system based on vision and arc cooperative sensing is used to control the height of the filling wire during the GTAW process and keep the GTAW in a continuous liquid bridge transition state to obtain better welding quality. It is characterized in that: the GTAW liquid bridge transition closed-loop control system based on vision and arc cooperative sensing is composed of a GTAW robot, a droplet transition image acquisition system, an arc signal acquisition system, a GTAW liquid bridge transition mode recognition system and a filling wire height control system; the GTAW welding robot is used to control the welding gun for welding; the droplet transition image acquisition system is composed of a high-speed camera and a high-speed camera guide rail and is used to shoot the GTAW droplet transition process; the arc signal acquisition system is used to collect the arc voltage signal between the wire end and the workpiece and the arc voltage signal between the tungsten electrode tip and the workpiece; the GTAW liquid bridge transition mode recognition system is used to identify the liquid bridge transition mode; the filling wire height control system is composed of a wire clamp and a clamp guide rail that can be used for the wire clamp to move up and down and is used to change the height of the filling wire so that the liquid bridge transition mode is in a continuous liquid bridge transition mode. The GTAW liquid bridge transition closed-loop control system based on vision and arc cooperative sensing is shown in the figure below. Figure 1 As shown; the structure diagram of the droplet transfer image acquisition system and the wire filling height control system is shown in Figure 2 shown.

[0004] The GTAW liquid bridge transition closed-loop control system based on vision and arc collaborative sensing is characterized by: using a GTAW robot to perform welding in a liquid bridge transition mode; the liquid bridge transition mode is divided into a continuous liquid bridge transition mode, an intermittent liquid bridge transition mode and a short-circuit liquid bridge transition mode; the droplet transition state is divided into three transition states, namely a, b and c; the state a is a state in which the droplet is not in contact with the molten pool; the state b is a state in which the droplet is just in contact with the molten pool; the state c is a state in which the welding wire is completely extended into the molten pool; the droplet transition state in the continuous liquid bridge transition mode is always maintained in state b; the droplet transition state in the intermittent liquid bridge transition mode is irregularly exchanged between states a and b; the droplet transition state in the short-circuit liquid bridge transition mode is always maintained in state c. The schematic diagram of droplet transition state a is shown in the figure Figure 4 As shown; the schematic diagram of the droplet transition state b is as follows Figure 5 As shown; the schematic diagram of the droplet transition state c is as follows Figure 6 shown.

[0005] The GTAW liquid bridge transition closed-loop control system based on vision and arc collaborative sensing is characterized in that: a molten droplet transition image acquisition system is used to collect image information of the molten droplet transition process; the molten droplet transition image acquisition system is composed of a high-speed camera and a high-speed camera guide rail; the high-speed camera is equipped with a filter that can filter arc light and can move up and down the high-speed camera guide rail; after welding starts, the high-speed camera moves to a position where the high-speed camera frame can completely include the tungsten electrode tip, the melting end of the welding wire, the molten pool and the molten droplet, and the focal length of the high-speed camera is adjusted to clearly capture the shape of the molten droplet.

[0006] The GTAW liquid bridge transition closed-loop control system based on vision and arc cooperative sensing is characterized in that: the arc signal information of the droplet transition process is collected by the arc signal acquisition system; after the welding starts, the arc signal acquisition system collects the arc voltage from the tungsten electrode tip to the workpiece and the arc voltage from the welding wire end to the workpiece over time; the arc voltage from the tungsten electrode tip to the workpiece is proportional to the distance from the tungsten electrode tip to the workpiece. 钨 = proportional relationship of kh; U 钨 is the arc voltage from the tip of the tungsten electrode to the workpiece; k is the proportional coefficient; h is the distance from the tip of the tungsten electrode to the workpiece; the arc voltage from the end of the welding wire to the workpiece has different images under different liquid bridge transition modes.

[0007] The GTAW liquid bridge transition closed-loop control system based on vision and arc cooperative sensing is characterized in that: the liquid bridge transition mode is identified by using the GTAW liquid bridge transition mode recognition system; the liquid bridge transition recognition system divides the continuous liquid bridge transition mode, the intermittent liquid bridge transition mode and the short-circuit liquid bridge transition mode into mode 1, mode 2 and mode 3; the liquid bridge transition mode recognition system uses time r as a cycle; within the time r, the number of images transmitted by the droplet transition image acquisition system is n; the arc signal acquisition system transmits the arc signal feature image within the time r; before welding starts, the liquid bridge transition mode recognition system refers to typical examples of the three liquid bridge transition modes and calculates the invariant matrices of the three liquid bridge transition modes and records the arc signal characteristics of the three liquid bridge transition modes; after welding starts, the liquid bridge transition mode recognition system receives the image information transmitted from the droplet transition image acquisition system and the arc signal information transmitted by the arc signal acquisition system; the liquid bridge transition mode recognition system confirms the n images to be confirmed area; the size of the area to be confirmed is α×βpx 2; The liquid bridge transition mode recognition system calculates 7 Hu invariant moments of n images respectively and compares them with the invariant moments of mode 1 calculated under the typical example to obtain n similarities, and calculates the average similarity X; when X>T, the image similarity is judged to be consistent with liquid bridge transition mode 1; A is the lowest similarity; when X<T, the 7 Hu invariant moments of n images are calculated and compared with the invariant moments of mode 2 calculated under the typical example to obtain n similarities, and calculate the average similarity Y; when Y>T, the image similarity is judged to be consistent with liquid bridge transition mode 2; when Y<T, the 7 Hu invariant moments of n images are calculated and compared with the invariant moments of mode 3 calculated under the typical example to obtain n similarities, and calculate the average similarity Z; when Z>T, the image similarity is judged to be consistent with liquid bridge transition mode 3; when Z<T, it is judged not to belong to any liquid bridge transition mode; after the liquid bridge transition mode recognition system makes a preliminary judgment through the image, it uses a dynamic time warping algorithm to compare the actual arc signal within time r with the image in the typical example 1 The arc signal characteristics obtained under the typical example 2 are compared to obtain the regular path distance x; when x<t, when the image similarity is judged as mode 1, the liquid bridge transition mode is judged to be mode 1, and when the image similarity is judged not to be mode 1, it is judged not to belong to any liquid bridge transition; t is the maximum allowable regular path distance; when x>t, the dynamic time warping algorithm is used to compare the actual arc signal in time r with the arc signal characteristics obtained under typical example 2 to obtain the regular path distance y; when y<t, when the image similarity is judged as mode 2, the liquid bridge transition mode is judged to be mode 2, and when the image similarity is judged not to be mode 2, it is judged not to belong to any liquid bridge transition; when y>t, the dynamic time warping algorithm is used to compare the actual arc signal in time r with the arc signal characteristics obtained under typical example 3 to obtain the regular path distance z; when z<t, when the image similarity is judged as mode 3, the liquid bridge transition mode is judged to be mode 3, and when the image similarity is judged not to be mode 3, it is judged not to belong to any liquid bridge transition; when z>t, it is judged not to belong to any liquid bridge transition mode. Schematic diagram of the area to be confirmed in the liquid bridge transition image acquisition system Figure 3 As shown; Based on the Hu unchanged liquid bridge mode judgment flow chart as shown Figure 7 As shown; the flow chart of liquid bridge mode judgment based on arc signal characteristics is as follows Figure 8 shown.

[0008] The GTAW liquid bridge transition closed-loop control system based on vision and arc collaborative sensing is characterized in that: the filler wire height is controlled by a filler wire height control system to maintain the liquid bridge transition mode in an efficient continuous liquid bridge transition mode; the control method is that when the liquid bridge transition mode is a short-circuit liquid bridge transition mode, the filler wire height control system controls the tip of the welding wire to move upward by Hmm; when the liquid bridge transition mode is an intermittent liquid bridge transition, the filler wire height control system controls the tip of the welding wire to move downward by Hmm; when the liquid bridge transition mode is a continuous liquid bridge transition, the filler wire height control system controls the tip of the welding wire to maintain a constant height; the moment when the filler wire height control system controls the tip of the welding wire to move is R seconds after the filler wire height control system receives the corresponding liquid bridge transition mode from the liquid bridge transition mode recognition system each time.

[0009] Advantageous Effects of the Invention

[0010] The present invention relates to the field of GTAW filling wire control, and is a GTAW liquid bridge transition closed-loop control system based on vision and arc collaborative sensing. In order to maintain a continuous liquid bridge transition mode for welding during the GTAW process, a GTAW liquid bridge transition closed-loop control system based on vision and arc collaborative sensing is proposed. A GTAW robot is used to control the welding action; a droplet transition image acquisition system is used to photograph the GTAW droplet transition process; an arc signal acquisition system is used to collect arc voltage signals between the welding wire end and the workpiece and arc voltage signals between the tungsten electrode tip and the workpiece; a GTAW liquid bridge transition mode recognition system is used to identify the liquid bridge transition mode; and a filling wire height control system is used to make the liquid bridge transition mode in a continuous liquid bridge transition mode. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Figure 1 Diagram of the GTAW liquid bridge transition closed-loop control system based on vision and arc collaborative sensing.

[0012] In the figure, 1 is a control box, 2 is a control bus, 3 is a communication bus, 4 is an engineering machine, 5 is a GTAW robot forward and backward moving guide rail, 6 is welding wire, 7 is a filler wire height control system, 8 is a welding gun, 9 is a GTAW robot up and down moving guide rail, 10 is a GTAW robot left and right moving guide rail, 11 is a high-speed camera guide rail, 12 is a high-speed camera, 13 is a workpiece, 14 is a working platform, 15 is a ground wire, and 16 is a collection bus

[0013] Figure 2 This is the structural diagram of the droplet transfer image acquisition system and the wire filling height control system.

[0014] In the figure, 1 is a high-speed camera guide rail, 2 is a high-speed camera, 3 is a welding wire clamp, 4 is a clamp guide rail, and 5 is a welding gun.

[0015] Figure 3 Schematic diagram of the area to be confirmed in the liquid bridge transition image acquisition system.

[0016] Figure 4 Schematic diagram of the droplet transition state a.

[0017] Figure 5 Schematic diagram of the droplet transition state b.

[0018] Figure 6 Schematic diagram of the droplet transition state c.

[0019] Figure 7 This is a flow chart based on the judgment of Hu unchanged liquid bridge mode.

[0020] Figure 8 The figure is a flow chart for determining the liquid bridge mode based on arc signal characteristics. DETAILED DESCRIPTION

[0021] In order to better express the technical solution and beneficial effects of the present invention, the present invention is further described in detail below with reference to the accompanying drawings and implementation examples. The implementation method of the present invention is not limited thereto.

[0022] Step 1: Liquid bridge transition mode identification.

[0023] In the GTAW welding droplet transition process, the liquid bridge transition mode needs to be identified. If the liquid bridge transition mode is not identified, the liquid bridge transition mode cannot be controlled to remain in the continuous liquid bridge transition mode. In view of this problem, the present invention discloses a GTAW liquid bridge transition closed-loop control system based on vision and arc collaborative sensing. After the welding starts, the droplet transition image acquisition system controls the high-speed camera to completely capture the droplet transition process image and transmit it to the GTAW liquid bridge transition mode recognition system. The arc signal acquisition system collects the arc signal information of the droplet transition process and transmits it to the GTAW liquid bridge transition mode recognition system. After receiving the signal, the GTAW liquid bridge transition mode recognition system identifies the liquid bridge transition mode and transmits the result to the filler wire height control system.

[0024] Step 2: Control the liquid bridge transition mode to maintain the continuous liquid bridge transition mode.

[0025] In the GTAW welding droplet transition process, the height of the filler wire needs to be controlled in real time. If the height of the filler wire is too high, the liquid bridge transition mode will be in the discontinuous liquid bridge transition mode; if the height of the filler wire is too low, the liquid bridge transition mode will be in the short-circuit liquid bridge transition mode. In view of this problem, the present invention discloses a GTAW liquid bridge transition closed-loop control system based on vision and arc collaborative sensing. After the welding starts, the filler wire height control system receives a signal from the GTAW liquid bridge transition mode recognition system and adjusts the filler wire height. When the liquid bridge transition mode is the short-circuit liquid bridge transition mode, the filler wire height control system controls the tip of the welding wire to move upward by Hmm; when the liquid bridge transition mode is the discontinuous liquid bridge transition, the filler wire height control system controls the tip of the welding wire to move downward by Hmm; when the liquid bridge transition mode is the continuous liquid bridge transition, the filler wire height control system controls the tip of the welding wire to maintain a constant height.

Claims

1. The GTAW liquid bridge transition closed-loop control system based on vision and arc cooperative sensing is used to control the wire height during the wire-filling GTAW process and keep the GTAW in a continuous liquid bridge transition state to obtain better welding quality. It is characterized by: The GTAW liquid bridge transition closed-loop control system based on vision and arc collaborative sensing is composed of a GTAW robot, a molten droplet transition image acquisition system, an arc signal acquisition system, a GTAW liquid bridge transition mode recognition system and a filler wire height control system; the GTAW welding robot is used to control the welding gun for welding; the molten droplet transition image acquisition system is composed of a high-speed camera and a high-speed camera guide rail and is used to photograph the GTAW molten droplet transition process; the arc signal acquisition system is used to collect the arc voltage signal between the welding wire end and the workpiece and the arc voltage signal between the tungsten electrode tip and the workpiece; the GTAW liquid bridge transition mode recognition system is used to identify the liquid bridge transition mode; the filler wire height control system is composed of a welding wire clamp and a clamp guide rail for the welding wire clamp to move up and down, and is used to change the filler wire height so that the liquid bridge transition mode is in a continuous liquid bridge transition mode.

2. The GTAW liquid bridge transition closed-loop control system based on vision and arc cooperative sensing according to claim 1 is characterized in that: A GTAW robot is used to perform welding in a continuous liquid bridge transition mode; the liquid bridge transition mode is divided into a continuous liquid bridge transition mode, an intermittent liquid bridge transition mode and a short-circuit liquid bridge transition mode; the molten droplet transition state is divided into three transition states, namely a, b and c; the state a is a state in which the molten droplet is not in contact with the molten pool; the state b is a state in which the molten droplet is just in contact with the molten pool; the state c is a state in which the welding wire is completely extended into the molten pool; the molten droplet transition state in the continuous liquid bridge transition mode is always maintained in state b; the molten droplet transition state in the intermittent liquid bridge transition mode is irregularly exchanged between states a and b; the molten droplet transition state in the short-circuit liquid bridge transition mode is always maintained in state c.

3. The GTAW liquid bridge transition closed-loop control system based on vision and arc cooperative sensing according to claim 1 is characterized in that: A molten droplet transfer image acquisition system is used to collect image information of the molten droplet transfer process; the molten droplet transfer image acquisition system is composed of a high-speed camera and a high-speed camera guide rail; the high-speed camera is equipped with a filter that can filter arc light and can move up and down the high-speed camera guide rail; after welding starts, the high-speed camera moves to a position where the high-speed camera frame can completely include the tungsten electrode tip, the melting end of the welding wire, the molten pool and the molten drop, and the focal length of the high-speed camera is adjusted to clearly capture the shape of the molten droplet.

4. The GTAW liquid bridge transition closed-loop control system based on vision and arc cooperative sensing according to claim 1 is characterized in that: The arc signal acquisition system is used to collect arc signal information of the droplet transfer process; after the welding starts, the arc signal acquisition system collects the arc voltage from the tungsten electrode tip to the workpiece and the arc voltage from the welding wire end to the workpiece over time; the arc voltage from the tungsten electrode tip to the workpiece is proportional to the distance from the tungsten electrode tip to the workpiece. 钨 = proportional relationship of kh; U 钨 is the arc voltage from the tip of the tungsten electrode to the workpiece; k is the proportional coefficient; h is the distance from the tip of the tungsten electrode to the workpiece; the arc voltage from the end of the welding wire to the workpiece has different images under different liquid bridge transition modes.

5. The GTAW liquid bridge transition closed-loop control system based on vision and arc cooperative sensing according to claim 1 is characterized in that: The GTAW liquid bridge transition mode recognition system is used to identify the liquid bridge transition mode; the liquid bridge transition recognition system divides the continuous liquid bridge transition mode, the intermittent liquid bridge transition mode and the short-circuit liquid bridge transition mode into mode 1, mode 2 and mode 3; the liquid bridge transition mode recognition system uses time r as a cycle; within the time r, the number of images transmitted by the droplet transition image acquisition system is n; the arc signal acquisition system transmits the arc signal feature image within the time r; before welding starts, the liquid bridge transition mode recognition system refers to typical examples of the three liquid bridge transition modes and calculates the invariant matrices of the three liquid bridge transition modes and records the arc signal features of the three liquid bridge transition modes; after welding starts, the liquid bridge transition mode recognition system receives the image information transmitted from the droplet transition image acquisition system and the arc signal information transmitted by the arc signal acquisition system; the liquid bridge transition mode recognition system confirms the area to be confirmed of n images; the size of the area to be confirmed is α×βpx 2 ; The liquid bridge transition mode recognition system calculates the 7 Hu invariant moments of n images respectively, compares them with the invariant moments of mode 1 obtained under typical examples to get n similarity degrees, and calculates the average similarity degree X; when X>T, the image similarity degree judgment conforms to the liquid bridge transition mode 1; A is the lowest similarity degree; when X<T, calculate the 7 Hu invariant moments of n images, compare them with the invariant moments of mode 2 obtained under typical examples to get n similarity degrees, and calculate the average similarity degree Y; when Y>T, the image similarity degree judgment conforms to the liquid bridge transition mode 2; when Y<T, calculate the 7 Hu invariant moments of n images, compare them with the invariant moments of mode 3 obtained under typical examples to get n similarity degrees, and calculate the average similarity degree Z; when Z>T, the image similarity degree judgment conforms to the liquid bridge transition mode 3; when Z<T, it is judged not to belong to any liquid bridge transition mode; after the preliminary judgment by the liquid bridge transition mode recognition system through the image, the dynamic time warping algorithm is used to compare the actual arc signal within time r with the arc signal characteristics obtained under typical example 1 to get the warping path distance x; when x<t, if the image similarity degree judgment is mode 1, judge that the liquid bridge transition mode is mode 1, if the image similarity degree judgment is not mode 1, judge not to belong to any liquid bridge transition; t is the maximum allowable warping path distance; when x>t, use the dynamic time warping algorithm to compare the actual arc signal within time r with the arc signal characteristics obtained under typical example 2 to get the warping path distance y; when y<t, if the image similarity degree judgment is mode 2, judge that the liquid bridge transition mode is mode 2, if the image similarity degree judgment is not mode 2, judge not to belong to any liquid bridge transition; when y>t, use the dynamic time warping algorithm to compare the actual arc signal within time r with the arc signal characteristics obtained under typical example 3 to get the warping path distance z; when z<t, if the image similarity degree judgment is mode 3, judge that the liquid bridge transition mode is mode 3, if the image similarity degree judgment is not mode 3, judge not to belong to any liquid bridge transition; when z>t, judge not to belong to any liquid bridge transition mode.

6. The GTAW liquid bridge transition closed-loop control system based on vision and arc cooperative sensing according to claim 1 is characterized in that: The filler wire height control system is used to control the filler wire height so that the liquid bridge transition mode is maintained in an efficient continuous liquid bridge transition mode; the control method is that when the liquid bridge transition mode is a short-circuit liquid bridge transition mode, the filler wire height control system controls the tip of the welding wire to move upward by Hmm; when the liquid bridge transition mode is an intermittent liquid bridge transition, the filler wire height control system controls the tip of the welding wire to move downward by Hmm; when the liquid bridge transition mode is a continuous liquid bridge transition, the filler wire height control system controls the tip of the welding wire to maintain a constant height; the moment when the filler wire height control system controls the tip of the welding wire to move is R seconds after the filler wire height control system receives the corresponding liquid bridge transition mode from the liquid bridge transition mode recognition system each time. .

Images