Multivariable cooperative control system and method for regulating and controlling size of molten pool in real time

Abstract

The invention relates to a multivariable cooperative control system and method for regulating and controlling the size of a molten pool in real time, the multivariable cooperative control system comprises an electric arc additive manufacturing system, a binocular passive vision sensing system and a double-fuzzy PID control system, and the electric arc additive manufacturing system is used for stacking metal thin-wall parts according to a preset path according to initial process parameters given by a welding machine; the binocular passive vision sensing system processes the molten pool image through the upper computer and then extracts corresponding molten pool width and molten pool height detection values in real time; and the double-fuzzy PID control system is used for adjusting the control variable of the width of the molten pool and the control variable of the height of the molten pool in real time according to a fuzzy control rule when the deviation value between the detected values of the width of the molten pool and the height of the molten pool and a preset value exceeds a threshold value. Closed-loop feedback control can be carried out on the width and height of the molten pool at the same time, the error variation between the detected value and the preset value of the size of the molten pool is kept within a constant range, and the forming stability of the metal thin-wall part is improved.

Description

Technical Field

[0001] The present invention relates to a system and method for regulating the size of a molten pool, and in particular to a multivariable collaborative control system and method for regulating the size of a molten pool in real time. Background Art

[0002] Additive manufacturing is a technology based on the discrete-deposition principle, which converts the CAD model of a structural part into a physical object by slicing it layer by layer and then stacking it layer by layer. Among them, arc additive manufacturing (DED-Arc) is an important branch of metal additive manufacturing technology. Compared with other metal additive manufacturing technologies such as laser additive manufacturing technology and electron beam additive manufacturing technology, this technology has the advantages of low production cost, high forming efficiency, high material utilization, short production cycle, and high part structural strength. It has a wide range of applications and has great advantages for the forming of large-sized complex metal structural parts. In addition, gas metal arc (GMA) additive manufacturing is an important branch of WAAM technology. It can continuously melt the metal wire fed into the arc through the gas metal arc, and stack metal structural parts layer by layer from bottom to top. It has the advantages of high deposition efficiency and relatively stable position of the molten droplet entering the molten pool. At present, this technology has been widely used in vehicles, rail transit, industrial mold manufacturing, aerospace, military industry and other scenarios.

[0003] However, in the process of arc additive manufacturing, due to the instability and fluctuation of process parameters such as current, voltage, scanning speed, the substrate is not preheated, and the substrate surface is severely oxidized, the forming size of the formed metal structure is often unable to keep consistent with the set value of the layered slice, and the more complex the structure is, the more serious this situation is. In addition, with the increase of the number of cladding layers, the heat dissipation conditions of each layer of the metal structure gradually deteriorate, resulting in excessive interlayer temperature and serious heat accumulation, which causes defects such as flow and hump in the molten pool, and ultimately causes the surface of the structure to be rough and uneven, the forming morphology is poor, the molten pool size accuracy is low, and residual stress is generated, which seriously affects the application of arc additive manufacturing technology in the high-efficiency and low-cost manufacturing of large and complex structural parts. Therefore, how to effectively dynamically control the heat input of each cladding layer so that the molten pool size (molten pool width and molten pool height) in the arc additive manufacturing process is always consistent with the set value of the layered slice, thereby improving the forming accuracy and improving the surface morphology and quality is the key to solving the above problems.

[0004] At present, the methods for controlling the size of the molten pool can be preliminarily divided into offline control and online control. Among them, offline control can establish a mathematical model of its stacking size and process parameters according to the effect of thin-walled parts forming, and artificially adjust the process parameters according to the model to obtain metal thin-walled parts with better morphology. However, this method is too dependent on manual experience, and due to the lack of real-time monitoring of the stacking size, there may be a large error between the established mathematical model and the actual situation. In addition, the control methods for the molten pool size proposed at present mainly focus on the control of a single variable (molten width or molten height), and there are few studies on the simultaneous control of molten width and molten height. There is still a lot of room for development in this field. Summary of the invention

[0005] Purpose of the invention: The purpose of the present invention is to propose a multivariable collaborative control system and method for real-time regulation of the molten pool size, which can simultaneously perform closed-loop feedback control on the molten pool width and the molten pool height, so that the error change between the detected value of the molten pool size and the preset value is maintained within a constant range, thereby improving the stability of the forming of thin-walled metal parts.

[0006] Technical solution: The present invention includes an arc additive manufacturing system, a binocular passive vision sensor system and a dual fuzzy PID control system, wherein the arc additive manufacturing system stacks metal thin-walled parts according to a preset path based on the initial process parameters given by the welding machine; the binocular passive vision sensor system processes the molten pool image through a host computer and extracts the corresponding molten pool width and molten pool height detection values ​​in real time; the dual fuzzy PID control system is used to adjust the control variables of the molten pool width and the control variables of the molten pool height in real time according to the fuzzy control rules when the deviation between the detection values ​​of the molten pool width and the molten pool height and the preset values ​​exceeds a threshold.

[0007] The arc additive manufacturing system includes an additive manufacturing platform, a robot teaching pendant and a control box. A base plate is provided on the top of the additive manufacturing platform, a welding gun is provided above the base plate, the welding gun is connected to a robotic arm and a wire feeder, and the robotic arm is connected to the welding machine.

[0008] The binocular passive visual sensing system includes a width industrial camera and a height industrial camera, and both the width industrial camera and the height industrial camera are connected to a host computer to perform image processing on the molten pool image taken by the camera.

[0009] The height industrial camera is clamped at the front end of the welding gun and fixed horizontally in front of the metal thin-walled part, and is used to capture the molten height image during the arc additive manufacturing molten pool forming process; the width industrial camera is clamped at the side of the welding gun, and is used to capture the molten width image during the arc additive manufacturing molten pool forming process.

[0010] The start-up time of the width industrial camera and the height industrial camera is: when the welding gun moves 8-10 mm, the camera is turned on and the molten pool image is detected to eliminate the instability at the arc starting and arc extinguishing points of the thin-walled metal parts.

[0011] The dual fuzzy PID control system includes a fuzzy PID controller for molten pool width and a fuzzy PID controller for molten pool height. The fuzzy PID controller for molten pool width controls the molten pool width by adjusting the current change, and the fuzzy PID controller for molten pool height controls the molten pool height by adjusting the voltage change.

[0012] The membership function of the input variable of the double fuzzy PID controller is a double Gaussian distribution function, and the membership function of the output variable is a triangular function.

[0013] A welding fume purifier is provided on one side of the welding gun, and a protective gas device is provided on the other side. The welding fume purifier is used to absorb welding fume generated during the accumulation process, and the protective gas device is used to ensure that the welding wire can be melted and formed.

[0014] A thermocouple is installed on the surface of the substrate to detect the temperature of the molten pool.

[0015] A multivariable collaborative control method for real-time regulation of molten pool size comprises the following steps:

[0016] Install and debug various components for additive manufacturing of thin-walled metal parts;

[0017] Take an image of the molten pool deposition process during arc additive manufacturing, and obtain the molten pool width detection value W through image processing algorithm m And the molten pool height detection value H m , the weld width detection value W m And melting height detection value H m Respectively with the preset value of weld width W set And the melting height preset value H set The difference is used to obtain the weld width deviation ΔW and the weld height deviation ΔH;

[0018] When the errors between the actual values ​​of the molten pool width and the molten pool height and the preset values ​​and their error change rates exceed the threshold range, the controller parameters are adjusted in real time according to the fuzzy control rules, and the control objects Δi and Δu of the molten width controller and the molten height controller are changed in real time, so that the stacking width deviation ΔW is less than 0.5mm and the stacking height deviation ΔH is less than 0.3mm.

[0019] Beneficial effects: The present invention has the following advantages:

[0020] (1) The present invention designs two controllers for adjusting the molten pool width and the molten pool height respectively. The width controller controls the molten pool width by adjusting the current change, and the height controller controls the molten pool height by adjusting the voltage change. The two controllers do not interfere with each other and do not affect each other. The molten pool size is controlled by dual variables, thereby effectively improving the problem of poor morphology of metal thin-walled parts, and finally producing low-carbon steel metal thin-walled parts with smooth surface, good mechanical properties and microstructure;

[0021] (2) The present invention can realize the coordinated control of the molten pool width and the molten pool height through a binocular passive visual sensing system, mainly through a width visual sensor and a height visual sensor to realize the simultaneous detection of the molten pool width and the molten pool height, and through image processing technology to realize the real-time extraction of the molten pool size;

[0022] (3) During the molten pool forming process, when the deviation between the detected value and the actual value of the molten pool size exceeds the threshold range, the parameters of the dual fuzzy PID controller can be adjusted in real time according to the fuzzy control rules formulated based on the previous process exploration experiments and expert experience, thereby adjusting the control object Δi of the melt width controller and the control object Δu of the melt height controller in real time, so that the molten pool size is stabilized near the preset value. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 A block diagram of a multivariable collaborative control system for real-time regulation of molten pool size according to the present invention;

[0024] Figure 2 It is a schematic structural diagram of the arc additive manufacturing system in the present invention;

[0025] Figure 3 Schematic diagram of the structure of the binocular passive visual sensing system in the present invention;

[0026] Figure 4 It is a schematic diagram of a step response experiment for solving a transfer function in a double fuzzy PID control system of the present invention;

[0027] Figure 5 It is the structural diagram of the double fuzzy PID control system in the present invention;

[0028] Figure 6 It is a visualization diagram of the dynamic relationship between input and output variables of the double fuzzy PID control system in the present invention;

[0029] Figure 7 A flow chart of the multivariable collaborative control method for real-time regulation of molten pool size provided by the present invention. DETAILED DESCRIPTION

[0030] The present invention will be further described below in conjunction with the accompanying drawings.

[0031] Example 1

[0032] like Figure 1As shown, the multivariable collaborative control system for real-time control of the molten pool size of this embodiment includes an arc additive manufacturing system, a binocular passive visual sensor system, and a dual fuzzy PID control system. The arc additive manufacturing system stacks thin-walled metal parts according to a preset path based on the initial process parameters given by the welding machine; the process parameters include welding current, welding voltage, scanning speed, interlayer cooling time, shielding gas flow, welding gun lifting height, and dry extension length; the binocular passive visual sensor system can extract the corresponding molten pool width and molten pool height detection values ​​in real time after processing the molten pool image by the host computer; the dual fuzzy PID control system is used to adjust the control variables of the molten pool width and the control variables of the molten pool height in real time according to the fuzzy control rules when the deviation between the detection values ​​of the molten pool width and the molten pool height and the preset values ​​exceeds the threshold value, so as to realize real-time control of the molten pool size.

[0033] like Figure 2 As shown, the arc additive manufacturing system includes an additive manufacturing platform 3, a robot teaching pendant 13 and a control box 16. A substrate 5 is provided on the top of the additive manufacturing platform 3. The substrate 5 is a low-carbon steel substrate. A thermocouple 4 is installed on the surface of the substrate 5 to detect the temperature of the molten pool. A welding gun 9 is provided above the substrate 5. The welding gun 9 is connected to a mechanical arm 10 and a wire feeder 11. The wire feeder 11 feeds the low-carbon steel welding wire to the bottom of the welding gun 9. The mechanical arm 10 is used to control the movement trajectory of the welding gun 9. The mechanical arm 10 is connected to a welding machine 14. The welding machine 14 is used to control the arcing and arcing of the welding gun 9. A welding smoke purifier 1 is provided on one side of the welding gun 9, and a protective gas device 15 is provided on the other side. The welding smoke purifier 1 is used to absorb the welding smoke generated during the accumulation process. The protective gas device 15 is used to ensure that the welding wire can be melted and formed. The teaching pendant 13 is used to control the operation of the welding machine 14, the welding gun 9 and the mechanical arm 10; the control box 16 is used to read and write the molten pool size control program.

[0034] The welding machine 14 can set the initial values ​​of process parameters such as welding current, welding voltage, scanning speed, interlayer cooling time, and shielding gas flow rate through the teaching pendant 13, and the wire feeding speed can be set by the wire feeder 11. The shielding gas device 15 needs to be turned on before the start of additive manufacturing and set the appropriate shielding gas flow rate through the teaching pendant 13 to prevent the low-carbon steel wire from being excessively oxidized by air during the additive manufacturing process, resulting in poor forming morphology or surface pores and other defects. The shielding gas is a mixed gas composed of 80% argon and 20% carbon dioxide.

[0035] like Figure 3As shown, the binocular passive visual sensing system includes a width industrial camera 8 for detecting the width of the molten pool and a height industrial camera 2 for detecting the height of the molten pool. Both the width industrial camera 8 and the height industrial camera 2 are connected to the host computer 12 to process the molten pool image taken by the camera. The height industrial camera 2 is clamped at the front end of the welding gun 9 by a clamp and fixed horizontally in front of the metal thin-walled part 6 to capture the molten height image during the arc additive manufacturing molten pool forming process; the width industrial camera 8 is clamped at the side of the welding gun 9 by a clamp to capture the molten width image during the arc additive manufacturing molten pool forming process. Both industrial cameras are equipped with a 680nm narrowband filter 7 to eliminate the interference caused by the strong arc light during the arc additive manufacturing process, and avoid large errors in the extraction of the molten pool size due to arc light interference. The image processing method includes: first, edge-preserving filtering is performed on the molten pool image through Open CV, and the area with large pixel differences at the edge of the molten pool image is retained, thereby completely retaining the edge information of the molten pool image. Secondly, the molten pool image is binarized, and the molten pool image is first corroded and then expanded to remove small spots and noise in the image. Finally, the contour of the molten pool image is extracted and the length is calculated to obtain real-time data of the molten pool size. In the specific implementation, the width industrial camera 8 and the height industrial camera 2 can be connected to one or more computers. Considering the instability of the arc starting and arc extinguishing points of thin-walled metal parts, the molten pool width industrial camera and the molten pool height industrial camera need to be turned on when the welding gun moves about 8-10mm after being connected to the host computer 12.

[0036] like Figure 4As shown in the figure, before designing the dual fuzzy PID controller, a step response experiment is required to study the dynamic characteristic relationship between the process parameters and the molten pool size according to the experimental equipment and experimental materials. Since the heat dissipation of each layer of cladding is different, the impact on the stacking size is different. In order to reduce this impact, the process parameters of the step response experiment will be stepped when the fifth layer is stacked, and the welding gun will be raised 1.5mm after each layer of printing is completed and then cooled. Among them, the weld bead length is 120mm, the step occurs at the midpoint of the weld bead, and the weld bead is taken as the y-axis, and the midpoint position is y=60mm. The process parameters for the first four layers of stacking are: stacking current 110A, stacking voltage 21V, and stacking speed 4mm / s. During the stacking process of the fifth layer, the process parameters of the fourth layer are kept unchanged. When the welding gun reaches the midpoint of the weld bead, that is, y=60mm, the process parameters such as stacking current, stacking voltage, and scanning speed are respectively subjected to single variable step response, and the camera sampling period is 0.04s. Due to the low temperature at the arc start of the weld and the insufficient wire feeding at the arc extinction, the weld is high at the arc start position and low at the arc extinction position, and the forming is unstable at the arc start and arc extinction. Therefore, when the visual sensing system is used to detect the molten pool forming process in real time, the 10mm outward from the arc start and the 10mm inward from the arc extinction will be excluded, that is, only the middle 100mm part of the weld except the arc start and arc extinction will be visually inspected and controlled in real time. Finally, according to the step response experimental results, two process parameters with the greatest influence on the molten pool size are selected and used as the control objects of the molten pool width controller and the molten pool height controller respectively.

[0037] like Figure 5 As shown, the dual fuzzy PID control system includes a fuzzy PID controller for the width of the molten pool and a fuzzy PID controller for the height of the molten pool. The fuzzy controller is used to regulate the error and the rate of change of the error between the detection value and the preset value of the molten pool size according to the fuzzy control rule table formulated by combining the previous process exploration experiments and expert experience. The parameters of the fuzzy controller are adjusted in real time and combined with the PID parameters to obtain the optimal parameters of the controller under the current conditions. The width controller controls the width of the molten pool by adjusting the change in current, and the height controller controls the height of the molten pool by adjusting the change in voltage. The two controllers do not interfere with each other and do not affect each other. The molten pool size is controlled collaboratively through dual variables. Among them, the molten pool width detection value W m and the preset value W set The error between w =W m -W set , error change rate Molten pool height detection value H m and the preset value H set The error between h =H m -H set , error change rate After the preliminary research and experiment on process parameters and dynamic characteristics of molten pool size, the molten pool width with good morphology of thin-walled metal parts was obtained, that is, the molten pool width was 7.2 mm and the molten pool height was 1.4 mm. Therefore, the preset value of the molten pool width in the controller was W. set =7.2mm, melt height preset value H set =1.4mm, on this basis, the error range between the detection value and the actual value of the molten pool width and molten pool height is allowed to be ±0.5mm and ±0.3mm respectively. Therefore, the threshold range of the molten pool width and molten pool height allowed by the dual fuzzy PID controller designed in this system is (7.2±0.5)mm and (1.4±0.3)mm respectively. When the detection value of the molten pool size exceeds the above threshold range, the dual fuzzy PID controller will adjust the controller parameters in real time according to the fuzzy control rules, thereby changing the values ​​of the molten width control object and the molten height control object, that is, the accumulation current change Δi and the accumulation voltage change Δu, so that the molten pool size can be stabilized in real time within the threshold range. The fuzzy control rules are compiled from {positive large, positive middle, positive small, zero, negative small, negative middle, negative large}, that is, {PB, PM, PS, ZO, NS, NM, NB} simple fuzzy language subsets. The fuzzy rules can realize the conversion of parameter values ​​from physical domain to fuzzy domain; the fuzzy rule tables specified according to the previous process exploration experiments and expert experience are shown in Tables 1 and 2 below. Where ΔK p , ΔK i , ΔK d They are the changes in the gains Kp, Ki, and Kd of the PID controller output by the fuzzy controller. The controller parameters obtained by combining the parameters of the PID controller in the current sampling period with the gain coefficient of the previous sampling period and the output value of the fuzzy controller are:

[0038]

[0039] Table 1 ΔK in the weld width controller p , ΔK i , ΔK d Fuzzy rule table

[0040]

[0041]

[0042] Table 2 ΔK in Melt Height Controller p , ΔK i , ΔK d Fuzzy rule table

[0043]

[0044] The membership function of the input variable of the dual fuzzy PID controller designed in this system is a double Gaussian distribution function, and the membership function of the output variable is a triangular function. The dynamic relationship diagram between the input variable and the output variable is as follows: Figure 6 shown.

[0045] In the specific implementation, both the binocular passive vision sensing system and the dual fuzzy PID control system need to be connected to the host computer, which can be connected to two computers respectively, or the visual detection system and the control system can be integrated into the software on one computer for centralized implementation.

[0046] In order to avoid the poor morphology of thin-walled metal parts due to oxidation of the substrate surface, the substrate surface needs to be polished with an angle grinder before the actual additive manufacturing begins. Among them, the models of low-carbon steel wire and substrate are ER50-6 and Q235B. The shielding gas flow rate is set to 15L / min. When it is lower than 15L / min, the thin-walled metal parts cannot be fully protected from oxidation interference by the external environment, and the lower shielding gas flow rate is easy to cause pore defects on the surface of the thin-walled metal parts; when it is higher than 15L / min, it is easy to cause waste of shielding gas and increase the loss of experimental materials. The initial process parameters of the welding machine can be set through the robot teaching pendant. When stacking single-pass multi-layer thin-walled parts, the welding current, welding voltage, scanning speed, interlayer cooling time, and wire dry extension length are set to 110A, 21V, 4mm / s, 5min and 1.5cm respectively. The scanning path of the welding gun is a reciprocating deposition, which can effectively avoid the situation where severe heat accumulation at the arc starting and arc extinguishing causes the collapse of thin-walled parts.

[0047] Example 2

[0048] like Figure 7 As shown, the multivariable collaborative control method for real-time regulation of the molten pool size of this embodiment includes the following steps:

[0049] Step 1: Install and fully debug the components for additive manufacturing of thin-walled metal parts.

[0050] In this embodiment, ER50-6 low carbon steel with a wire diameter of 1.2mm is selected as the welding wire of this system, and Q235B low carbon steel with a size of 200mm×200mm×10mm is selected as the substrate of this system. Before welding begins, an angle grinder is used to grind the surface of the substrate to remove the oxide layer. After pre-treatment of the substrate, it is fixed to the additive manufacturing platform with four bolts to ensure that the position of the substrate can be kept stable during the actual stacking process; after fixing the position of the substrate, the molten pool width industrial camera is installed on the side of the welding gun and the molten pool height industrial camera is installed at the front end of the welding gun through a special fixture and is horizontal to the weld. After installing the camera, it is connected to the host computer and the camera start time is set, that is, when the welding gun travels 8-10mm, the camera is turned on and the molten pool image is detected. After the ER50-6 low carbon steel welding wire is sent to the lower end of the welding gun through the wire feeder, the mixed shielding gas device composed of 80% argon and 20% concentration is turned on, and the shielding gas flow rate is set to 15L / min. At the same time, according to the previous research experiments on process parameters and dynamic characteristics of molten pool size, appropriate initial process parameters of the welding machine were set through the robot teaching pendant, and the molten pool size control program was written through the control box to stabilize the molten pool width and molten pool height within the threshold range of (7.2±0.5) mm and (1.4±0.3) mm, respectively, in preparation for the start of single-pass multi-layer metal thin-walled parts accumulation.

[0051] Step 2: Take an image of the molten pool deposition process during arc additive manufacturing, and obtain the molten pool width detection value W through image processing algorithm m And the molten pool height detection value H m , the weld width detection value W m And melting height detection value H m Respectively with the preset value of weld width W set And the melting height preset value H set The difference is used to obtain the weld width deviation ΔW and the weld height deviation ΔH.

[0052] The image processing steps are as follows: First, the edge-preserving filter is applied to the melt pool image through Open CV, and the areas with large pixel differences at the edge of the melt pool image are retained, thereby completely retaining the edge information of the melt pool image. Secondly, the melt pool image is binarized, and the melt pool image is first eroded and then expanded to remove small spots and noise in the image. Finally, the contour of the melt pool image is extracted and the length is calculated to obtain real-time data of the melt pool size.

[0053] Step 3: During the actual stacking process, when the errors between the actual values ​​of the molten pool width and the molten pool height and the preset values ​​and their error change rates exceed the threshold range, the controller parameters are adjusted in real time according to the fuzzy control rules, and the control objects Δi and Δu of the molten width controller and the molten height controller are changed in real time, so that the stacking width deviation ΔW is less than 0.5mm, and the stacking height deviation ΔH is less than 0.3mm.

Claims

1. A multivariable collaborative control system for real-time regulation of molten pool size, characterized in that: The invention comprises an arc additive manufacturing system, a binocular passive vision sensing system and a dual fuzzy PID control system, wherein the arc additive manufacturing system stacks metal thin-walled parts according to a preset path according to the initial process parameters given by the welding machine; the binocular passive vision sensing system processes the molten pool image through a host computer and extracts the corresponding molten pool width and molten pool height detection values ​​in real time; the dual fuzzy PID control system is used for adjusting the control variables of the molten pool width and the molten pool height in real time according to the fuzzy control rules when the deviation between the detection values ​​of the molten pool width and the molten pool height and the preset values ​​exceeds a threshold value.

2. A multivariable collaborative control system for real-time regulation of molten pool size according to claim 1, characterized in that: The arc additive manufacturing system includes an additive manufacturing platform, a robot teaching pendant and a control box. A base plate is provided on the top of the additive manufacturing platform, a welding gun is provided above the base plate, the welding gun is connected to a robotic arm and a wire feeder, and the robotic arm is connected to the welding machine.

3. A multivariable collaborative control system for real-time regulation of molten pool size according to claim 2, characterized in that: The binocular passive visual sensing system comprises a width industrial camera and a height industrial camera, and both the width industrial camera and the height industrial camera are connected to a host computer.

4. A multivariable collaborative control system for real-time regulation of molten pool size according to claim 3, characterized in that: The height industrial camera is clamped at the front end of the welding gun and is horizontally fixed in front of the metal thin-walled part; the width industrial camera is clamped at the side of the welding gun.

5. A multivariable collaborative control system for real-time regulation of molten pool size according to claim 4, characterized in that: The start time of the width industrial camera and the height industrial camera is: when the welding gun moves 8-10 mm, the camera is turned on and the molten pool image is detected.

6. A multivariable collaborative control system for real-time regulation of molten pool size according to claim 1, characterized in that: The dual fuzzy PID control system includes a fuzzy PID controller for molten pool width and a fuzzy PID controller for molten pool height. The fuzzy PID controller for molten pool width controls the molten pool width by adjusting the current change, and the fuzzy PID controller for molten pool height controls the molten pool height by adjusting the voltage change.

7. A multivariable collaborative control system for real-time regulation of molten pool size according to claim 6, characterized in that: The membership function of the input variable of the double fuzzy PID controller is a double Gaussian distribution function, and the membership function of the output variable is a triangular function.

8. A multivariable collaborative control system for real-time regulation of molten pool size according to claim 2, characterized in that: A welding fume purifier is arranged on one side of the welding gun, and a protective gas device is arranged on the other side.

9. A multivariable collaborative control system for real-time regulation of molten pool size according to claim 2, characterized in that: A thermocouple is installed on the surface of the substrate.

10. A multivariable collaborative control method for real-time regulation of molten pool size, characterized in that: The following steps are involved: Install and debug various components for additive manufacturing of thin-walled metal parts; Take an image of the molten pool deposition process during arc additive manufacturing, and obtain the molten pool width detection value W through image processing algorithm m And the molten pool height detection value H m , the weld width detection value W m And melting height detection value H m Respectively with the preset value of weld width W set And the melting height preset value H set The difference is used to obtain the weld width deviation ΔW and the weld height deviation ΔH; When the errors between the actual values ​​of the molten pool width and the molten pool height and the preset values ​​and their error change rates exceed the threshold range, the controller parameters are adjusted in real time according to the fuzzy control rules, and the control objects Δi and Δu of the molten width controller and the molten height controller are changed in real time, so that the stacking width deviation ΔW is less than 0.5mm and the stacking height deviation ΔH is less than 0.3mm.

Images