CN107427951A - Utilize the welding system of adaptive algorithm - Google Patents
Utilize the welding system of adaptive algorithm Download PDFInfo
- Publication number
- CN107427951A CN107427951A CN201680009100.8A CN201680009100A CN107427951A CN 107427951 A CN107427951 A CN 107427951A CN 201680009100 A CN201680009100 A CN 201680009100A CN 107427951 A CN107427951 A CN 107427951A
- Authority
- CN
- China
- Prior art keywords
- welding
- data
- actuator
- sensing
- actuator control
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B23—MACHINE TOOLS; METAL-WORKING NOT OTHERWISE PROVIDED FOR
- B23K—SOLDERING OR UNSOLDERING; WELDING; CLADDING OR PLATING BY SOLDERING OR WELDING; CUTTING BY APPLYING HEAT LOCALLY, e.g. FLAME CUTTING; WORKING BY LASER BEAM
- B23K9/00—Arc welding or cutting
- B23K9/32—Accessories
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B23—MACHINE TOOLS; METAL-WORKING NOT OTHERWISE PROVIDED FOR
- B23K—SOLDERING OR UNSOLDERING; WELDING; CLADDING OR PLATING BY SOLDERING OR WELDING; CUTTING BY APPLYING HEAT LOCALLY, e.g. FLAME CUTTING; WORKING BY LASER BEAM
- B23K9/00—Arc welding or cutting
- B23K9/095—Monitoring or automatic control of welding parameters
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B23—MACHINE TOOLS; METAL-WORKING NOT OTHERWISE PROVIDED FOR
- B23K—SOLDERING OR UNSOLDERING; WELDING; CLADDING OR PLATING BY SOLDERING OR WELDING; CUTTING BY APPLYING HEAT LOCALLY, e.g. FLAME CUTTING; WORKING BY LASER BEAM
- B23K9/00—Arc welding or cutting
- B23K9/095—Monitoring or automatic control of welding parameters
- B23K9/0953—Monitoring or automatic control of welding parameters using computing means
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B13/00—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion
- G05B13/02—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B13/00—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion
- G05B13/02—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric
- G05B13/0265—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric the criterion being a learning criterion
- G05B13/027—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric the criterion being a learning criterion using neural networks only
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B13/00—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion
- G05B13/02—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric
- G05B13/0265—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric the criterion being a learning criterion
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B13/00—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion
- G05B13/02—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric
- G05B13/0265—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric the criterion being a learning criterion
- G05B13/0275—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric the criterion being a learning criterion using fuzzy logic only
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B13/00—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion
- G05B13/02—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric
- G05B13/0265—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric the criterion being a learning criterion
- G05B13/0285—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric the criterion being a learning criterion using neural networks and fuzzy logic
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Software Systems (AREA)
- Plasma & Fusion (AREA)
- Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Medical Informatics (AREA)
- General Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Mechanical Engineering (AREA)
- Mathematical Physics (AREA)
- Fuzzy Systems (AREA)
- Theoretical Computer Science (AREA)
- Feedback Control In General (AREA)
- Manipulator (AREA)
Abstract
Welding system includes：Actuator (101) is welded, the welding actuator operates according to actuator control data；Sensing system (102), the sensing system are used to produce the sensing data relevant with weld seam；And processing system (103), the processing system are used for based on sensing data and represent that the welding data of material to be welded and welding condition determines actuator control data using adaptive algorithm.The welding system includes allowing users to input the training data interface (104,105) of actuator control data to train welding process.The processing system is configured to the actuator control data inputted according to welding data, via training data interface and the sensing data measured during welding process is trained to train adaptive algorithm.Therefore, the welding system can be trained such that user is in control loop.The adaptive algorithm can be based on such as artificial neural network.
Description
Technical field
The disclosure relates generally to automation, mechanization and/or robot welding.It is more specifically, this disclosure relates to a kind of
Welding system, it is related to a kind of method for training welding system, and is related to a kind of computer for being used to train welding system
Program.
Background technology
Welding is for by such as ship, railways means, building structure, boiler, pipeline, power station, aircraft, automobile etc.
The common method that the part of mechanical structure for good and all links.Although being welded with many benefits, however, by welding the link formed
Portion is often the most weak-strong test of mechanical structure.Especially, the defects of solder joint portion is problematic, because the defect is machine
The potential start-point of tool failure.As corollary, in most of situations, need to weld again when observing weld defect
Connect.Therefore, weld defect increase both cost and production time.In the company for making welding product, led by weld defect
The cost of cause is probably the important share of the turnover.If this share can be reduced, the direct benefit to industry will quite may be used
See.
In automation, mechanization and/or robot welding, welding system generally includes to weld actuator, for example, according to
The welding robot of actuator control data operation.In addition, welding system is generally included for producing the sensing relevant with weld seam
The sensing system of device data, and for based on sensing data and the welding data of instruction material to be welded and welding condition
To determine the processing system of actuator control data.Above-mentioned actuator control data includes such as welding current, welding
Voltage, speed of welding, the feed rate of welding wire and/or from contact nozzle to the welding parameter of the distance of soldered part.From
The distance of contact nozzle to part indicates the length of electric arc.Speed of welding is defined as the end of welding wire relative to soldered portion
Travel rate of the part on the direction of weld seam.Due to the nonlinear characteristic of welding process, it is difficult to develop for predicting welding process
Behavior model.Therefore, above-mentioned processing system is constructed to be permeable to determine welding parameter and possible other actuator controls
Data processed to realize that enough welding qualities are challenging under different welding conditions.
The content of the invention
The content of the invention of simplification presented below, to provide the basic comprehension of some aspects to various inventive embodiments.
Present invention is not the extensive overview ot of the present invention.It is neither intended to the key or important elements of the mark present invention, is not intended to
Delineate the scope of the present invention.Some designs of the present invention are only presented in the following content of the invention in simplified form, as the present invention's
The preamble in greater detail of exemplary embodiment.
According to the present invention, there is provided a kind of new welding system, the welding system include：
- welding actuator, the welding actuator operate according to actuator control data,
- sensing system, it is relevant with the weld seam as caused by the welding actuator that the sensing system is used for generation
Sensing data, such as Temperature Distribution,
- processing system, the processing system are used for based on the sensing data and instruction material to be welded and welding bar
The welding data of part determines the actuator control data using adaptive algorithm, and
- training data interface, the training data interface are used to enabling the user of the welding system to train
Welding process and at least a portion actuator control data inputted in the actuator control data.
The processing system is configured to hold according to the welding data, via described in the input of the training data interface
At least a portion actuator control data in row device control data and measured during the training welding process
The sensing data train the adaptive algorithm, so as to the adaptive algorithm is trained to can control with it is described
Train the corresponding welding process of welding process.Therefore, the welding system can be trained to so that as the welder that specialty is skilled
By via at least a portion actuator control data in actuator control data described in the training data interface setting come
During operation training welding process, the adaptive algorithm is under mode of learning.The adaptive algorithm can be based on such as people
Artificial neural networks " ANN ".The training data interface can be the part for operating the user interface of the welding system,
Or the training data interface can be single entity.
The welding system can also include accumulator system, the accumulator system be used to realize be used to storing with it is various
The database of the relevant event data of welding process.Such as instruction welding can be included by being related to the event data of given welding process
The welding data of main welding condition during material and welding process under consideration.In addition, event data can include
The sensing data recorded during actuator control data and welding process under consideration.Can also, welding system bag
Data interface is included, for being for example connected to the external memory system for realizing database via data transmission network.It is described
Welding system can also include being used for the movement for resetting welding actuator and for based on being recorded and given welding process
Relevant event data shows the device of sensing data so that people is then able to check the operation of welding system.
According to the present invention, a kind of new method for being used to train welding system is additionally provided, the welding system includes：
- welding actuator, the welding actuator operate according to actuator control data,
- sensing system, it is relevant with the weld seam as caused by the welding actuator that the sensing system is used for generation
Sensing data, and
- processing system, the processing system are used for based on the sensing data and instruction material to be welded and welding bar
The welding data of part determines the actuator control data using adaptive algorithm.
The method according to the invention includes：
- in order to train welding process, input the actuator via the training data interface of the welding system and control number
At least a portion actuator control data in, and
- according to the welding data, via the training data interface input the actuator control data in institute
The sensing data stated at least a portion actuator control data and measured during the training welding process comes
The adaptive algorithm is trained, trains welding process corresponding with described to be trained to the adaptive algorithm to control
Welding process.
According to the present invention, the new computer program of the welding system for training the above-mentioned type is additionally provided.According to this hair
Bright computer program includes computer executable instructions, and the computer executable instructions are used to control programmable processing system
With：
- in order to train welding process, receive the actuator control data from the training data interface of the welding system
In at least a portion actuator control data, and
Institute in the-actuator control data received according to the welding data, from the training data interface
The sensing data stated at least a portion actuator control data and measured during the training welding process comes
The adaptive algorithm of the welding system is trained, can be controlled and the training so as to which the adaptive algorithm is trained to
The corresponding welding process of welding process.
According to the present invention, a kind of new computer program product is additionally provided.The computer program product has including coding
According to the non-volatile computer-readable medium of the computer program of the present invention, such as CD " CD ".
The multiple exemplary and non-limiting example of the present invention is described in accompanying independent claim.
When read in conjunction with the accompanying drawings, by from the specific exemplary and following description of non-limiting example best
Understand not only on construction again on the various exemplary and non-limiting example of the invention of operating method and its in addition
Target and advantage.
Verb " comprising " and "comprising" are used as both being not excluded in the literature or do not require not record existing for feature
Opening limitation.Unless expressly stated otherwise, otherwise it is recited in mutually different dependent and is characterized in mutually freely combining
's.However, it should be understood that the "a" or "an" used through the literature, i.e. singulative, however not excluded that multiple.
Brief description of the drawings
Below in the sense that example and illustrate in greater detail with reference to the attached drawings the present invention it is exemplary and non-limiting
Embodiment and its advantage, wherein：
Fig. 1 shows the exemplary and schematic illustration of the welding system of non-limiting example according to the present invention, and
And
Fig. 2 shows the exemplary and method of non-limiting example according to the present invention for training welding system
Flow chart.
Embodiment
Fig. 1 shows the exemplary and schematic illustration of the welding system of non-limiting example according to the present invention.The weldering
Welding system includes the welding actuator 101 according to the operation of actuator control data.In the example scenario, actuator is welded
101 be welding robot, but the welding actuator also can be such as welding gantry (welding portal).In addition, should
Welding system includes sensing system 102, for producing the sensor relevant with the weld seam 110 as caused by welding actuator 101
Data.The welding system includes processing system 103, for based on sensing data and instruction material to be welded and welding condition
Welding data determines above-mentioned actuator control data using adaptive algorithm.Processing system 103 is communicably connected to welding and performed
Device 101.Above-mentioned welding condition can mean for example used welding technique, used welding gas (if yes),
The space such as physical dimension of weld groove 111, the thickness of welding wire 106 and/or welding position for weld seam.Welding technique can be
Such as melting welding, the melting welding can be, for example, that gas metal arc welding connects " GMAW ", such as Metallic Inert Gas " MIG " welds
Connect or metal active gas " MAG " welds, or Gas-Tungsten-Arc welding " GTAW ", also referred to as Wolfram Inert Gas " TIG " weld
Connect.The physical dimension of weld groove 111 can mean the width and depth of weld groove.In Fig. 1 in shown exemplary cases, in coordinate
Be 190 x directions on measure weld groove width and on the z directions of coordinate system 190 measure weld groove depth.Welding position limits
The direction of fixed weld seam 110 to be generated and the position for treating the part 107 and 108 by solder joint.Shown example in Fig. 1
In implementations, welding position make it that part is horizontal and weld seam is horizontal and is located on the upside of part.
Above-mentioned actuator control data includes welding parameter, such as welding current, weldingvoltage, speed of welding,
The feed rate of welding wire 106 and from contact nozzle to the distance for the part 107 and 108 for passing through solder joint.Welding current
It is the important welding parameter in electric arc welding, wherein welding wire burns the geometry of speed, fusion penetration and weld seam by welding electricity
Stream at least partially determines.Weldingvoltage is the potential difference between the surface of the weldpool in the end of welding wire and melting.Fusion area
Shape and weld reinforcement at least partially determined by weldingvoltage.In Fig. 1 in shown exemplary cases, speed of welding quilt
It is defined as travel rate of the end of welding wire on the y directions of coordinate system 190.In Fig. 1, speed of welding is described with arrow v.
In electric arc welding, the increase of speed of welding typically results in the reduction of the per unit length heat input of weld seam 110, welding wire burns speed
The reduction of rate and the reduction of weld reinforcement.The feed rate of welding wire 106 means that the spray nozzle part 109 of welding wire from welding actuator comes out
Speed.
Sensing system 102 can include such as thermograph sensor, for from caused weld seam 110 and from weldering
Stitch the area measure temperature on side.Measured temperature represents the hot temperature for how spreading to the region beside weld seam of especially description
Degree distribution.The Temperature Distribution is relevant with heat input speed so that high heat input speed causes broader Temperature Distribution, i.e. with compared with
Small heat input speed ratio, heat are more strongly spread.Sensing system 102 can include laser line scanning camera and for running
Image recognition is so as to the processor of the surface profile of weld seam caused by detecting.In Fig. 1 in shown exemplary cases, sitting
The surface profile of weld seam is measured in the xz planes of mark system 190.It is still possible that the sensing system is included produced by being used to measure
Weld seam surface profile non-contact distance detector.The operation of non-contact distance detector can be surveyed based on such as optical triangulation
Amount, wherein the light (such as laser beam) from the region reflection including weld seam is focused on photosensitive detector surface by optical lens
And position based on irradiation area on the photosensitive detector surface determines distance.
As mentioned earlier, weld actuator 101 to operate according to actuator control data, and processing system 103 is by structure
Cause to determine to perform using adaptive algorithm based on sensing data and the welding data of instruction material to be welded and welding condition
Device control data.The welding system includes training data interface, for enabling the user of the welding system to train
Welding process and at least a portion actuator control data inputted in actuator control data.Inputted via training data interface
Actuator control data can include the subset of such as welding parameter or welding parameter.Processing system 103 be configured to according to
With the relevant welding data of training welding process, via described in the actuator control data of training data interface input at least
A part of actuator control data and the sensing data that is measured during welding process is trained train adaptive algorithm,
So as to the adaptive algorithm is trained to can be controlled in the case of the interaction of no user with training welding process it is corresponding
Welding process.Therefore, the welding system can be trained to so that when specialty, skilled welder passes through via training data circle
It is adaptive to calculate when at least a portion actuator control data in face setting actuator control data trains welding process to run
Method is under mode of learning.Training data interface can be the part for operating the user interface 104 of welding system, or
Training data interface can be single entity.In the example scenario illustrated in Fig. 1, training data interface includes portable
Formula data input device 105, such as when user is near the part 107 and 108 by solder joint and user is used
When the welding helmet visually monitors training welding process, the portable data entry device 105 can be carried by user.
In the exemplary and welding system of non-limiting example according to the present invention, adaptive algorithm is manually neural
Network " ANN " realizes, the artificial neural network " ANN " can according to welding data, via the input of training data interface
At least a portion actuator control data in actuator control data and measured during welding process is trained
Sensing data learns.In the another exemplary and the welding system of non-limiting example according to the present invention, adaptively
Algorithm realized with fuzzy logic, the execution that the fuzzy logic can input according to welding data, via training data interface
At least a portion actuator control data in device control data and the sensing measured during welding process is trained
Device data are adapted.In the exemplary and welding system of non-limiting example according to the present invention, adaptive algorithm part
Ground employment artificial neural networks and partly realized with fuzzy logic.In the exemplary and non-limiting implementation according to the present invention
In the welding system of example, processing system is configured to support one group of adaptive algorithm, and for each welding process, based on weldering
Connect data and adaptive algorithm is selected among the adaptive algorithm group, the welding data passes through expression material for example to be welded, institute
The welding technique that uses, used welding gas (if yes), the physical dimension of space such as weld groove for weld seam, weldering
Weld task is specified in the thickness of silk and/or welding position.
In the example scenario illustrated in Fig. 1, the welding system is connected to data transmission network 112 so that
Processing system 103 is communicably connected to the data transmission network.Data transmission network 112 can be, for example, but be not necessarily ether
Net network.Accumulator system 113 is communicably connected to data transmission network 112, and has for storing with various welding processes
The database of the event data of pass is realized by means of accumulator system 113.The event data for being related to given welding process can be with
Welding data including for example specifying welding material and main welding condition, and actuator control data and under consideration
The sensing data recorded during welding process.The welding system can also include being used for the shifting for resetting welding actuator 101
Move and for showing the device of sensing data based on the event data relevant with given welding process recorded so that
User is then able to check the operation of welding system.Computer installation 114 including user interface is communicably connected to data biography
Defeated network 112.Computer installation 114 is advantageously able to read data from above-mentioned database and writes data into the data
Storehouse.Therefore, the event data of each welding process recorded can be provided with the quality for representing for example caused weld seam
The data of classification, the identifier of the user of welding system and/or the other data relevant with the welding process in consideration.
Fig. 2 shows the exemplary and method of non-limiting example according to the present invention for training welding system
Flow chart, the welding system include：
- welding actuator, the welding actuator operate according to actuator control data,
- sensing system, it is relevant with the weld seam as caused by the welding actuator that the sensing system is used for generation
Sensing data, and
- processing system, the processing system are used for based on the sensing data and instruction material to be welded and welding bar
The welding data of part determines the actuator control data using adaptive algorithm.
Methods described includes following action：
- action 201：In order to train welding process, via the training data interface input actuator control number of welding system
At least a portion actuator control data in, and
- action 202：According to welding data, via training data interface input actuator control data in described in extremely
Lack a part of actuator control data and the sensing data measured during welding process is trained to train adaptive calculation
Method, so as to the adaptive algorithm is trained to can be controlled in the case of the interaction of no user with training welding process it is relative
The welding process answered.
In the exemplary and method of non-limiting example according to the present invention, one during sensing data instruction is following
It is individual or multiple：From weld seam and from the area measure beside weld seam to temperature, and/or the surface profile of caused weld seam.
In the exemplary and method of non-limiting example according to the present invention, adaptive algorithm employment artificial neural networks
To realize, the artificial neural network can according to welding data, via training data interface input actuator control data
In at least a portion actuator control data and the sensing data that is measured during welding process is trained learn
Practise.
In the exemplary and method of non-limiting example according to the present invention, adaptive algorithm is with fuzzy logic come real
It is existing, the fuzzy logic can according to welding data, via described in the actuator control data of training data interface input
At least a portion actuator control data and the sensing data measured during welding process is trained are adapted.
In the exemplary and method of non-limiting example according to the present invention, actuator control data includes welding and joined
Number, the welding parameter include one or more of following：Welding current；Weldingvoltage；It is defined as the end phase of welding wire
For the speed of welding of travel rate of the soldered part on the direction of weld seam；The feed rate of welding wire；From contact nozzle
To the distance of soldered part.
In the exemplary and method of non-limiting example according to the present invention, in addition to indicating material to be welded, weldering
Connect data also indicate it is one or more of following：Welding technique to be used；Welding gas (if yes)；For weld seam
The physical dimension of space such as weld groove；The thickness of welding wire；Welding position.
Included according to the exemplary and non-limiting example method of the present invention：It is adaptive from one group based on welding data
Selection adaptive algorithm among algorithm.
Computer executable instructions, institute are included according to the exemplary and computer program of non-limiting example of the present invention
State computer executable instructions to be used to control programmable processing system, to perform and the above-mentioned example according to the present invention and non-limit
The relevant action of the method for any one in property embodiment processed.
Include being used to train welding system according to the exemplary and computer program of non-limiting example of the present invention
Software module, the welding system include：
- welding actuator, the welding actuator operate according to actuator control data,
- sensing system, it is relevant with the weld seam as caused by the welding actuator that the sensing system is used for generation
Sensing data, and
- processing system is may be programmed, the programmable processing system is used for be welded based on the sensing data and instruction
The welding data of material and welding condition determines the actuator control data using adaptive algorithm.
The software module includes computer executable instructions, and the computer executable instructions, which are used to controlling, described can compile
Journey processing system with：
- in order to train welding process, receive the actuator control data from the training data interface of the welding system
In at least a portion actuator control data, and
Institute in the-actuator control data received according to the welding data, from the training data interface
The sensing data stated at least a portion actuator control data and measured during the training welding process comes
The adaptive algorithm is trained, so as to which the adaptive algorithm to be trained to and can be controlled in the case of the interaction of no user
The welding process corresponding with the training welding process.
The software module can be for example with suitable programming language and with being suitable for the programming language and be suitable for
The subroutine or function that the compiler of programmable processing system in consideration is realized.It is worth noting that, with suitable programming language
Say that corresponding source code also illustrates that computer can perform software module, because source code package is containing the programmable processing system of control
The information needed for the action being presented above with performing of uniting and compiling only change the form of information.Furthermore, it is also possible that it can compile
Journey processing system is provided with interpretive program so that the source code realized with suitable programming language need not before runtime by
Compiling.
With good grounds hair of coding is included according to the exemplary and computer program product of non-limiting example of the present invention
The computer-readable medium of the computer program of bright exemplary embodiment, such as CD " CD ".
Carrying definition is encoded into according to the present invention's according to the exemplary and non-limiting example signal of the present invention
The information of the computer program of exemplary embodiment.
The specific example provided in description given above is not construed as limiting the model of appended claims
Enclose and/or applicability.Unless expressly stated otherwise, the list of the example otherwise provided in description given above and group
It is not exhaustive.
Claims (18)
1. a kind of welding system, including：
- welding actuator (101), the welding actuator (101) operate according to actuator control data,
- sensing system (102), the sensing system (102) are used to produce and the weld seam as caused by the welding actuator
Relevant sensing data, and
- processing system (103), the processing system (103) are used to utilize certainly based on the sensing data and welding data
Adaptive algorithm determines the actuator control data, and the welding data indicates material to be welded and welding condition,
Characterized in that, the welding system also includes training data interface (104,105), the training data interface is used to make
The control number of at least a portion actuator in the actuator control data can be inputted to train welding process by obtaining user
According to, and the processing system is configured to hold according to the welding data, via described in the input of the training data interface
At least a portion actuator control data in row device control data and measured during the training welding process
The sensing data train the adaptive algorithm, so as to the adaptive algorithm is trained to can control with it is described
Train the corresponding welding process of welding process.
2. welding system according to claim 1, wherein, the sensing system includes thermograph sensor, described
Thermograph sensor is used to refer to from the weld seam and from the area measure temperature beside the weld seam, the sensing data
Show measured temperature.
3. welding system according to claim 1 or 2, wherein, the sensing system include laser line scanning camera and
Processor, the processor are used for operation image and identified to detect the surface profile of caused weld seam, the sensor number
According to the surface profile of weld seam caused by instruction.
4. welding system according to claim 1 or 2, wherein, the sensing system includes non-contact distance detector,
The non-contact distance detector be used to measuring caused by weld seam surface profile, caused by sensing data instruction
The surface profile of weld seam.
5. the welding system according to any one of Claims 1-4, wherein, the adaptive algorithm is manually neural
Network realizes, the institute that the artificial neural network can input according to the welding data, via the training data interface
State at least a portion actuator control data in actuator control data and surveyed during the training welding process
The sensing data measured learns.
6. the welding system according to any one of claim 1 to 5, wherein, the adaptive algorithm fuzzy logic
To realize, the fuzzy logic can according to the welding data, via the training data interface input the actuator
At least a portion actuator control data in control data and the institute measured during the training welding process
Sensing data is stated to be adapted.
7. the welding system according to any one of claim 1 to 6, wherein, the actuator control data includes weldering
Parameter is connect, the welding parameter includes one or more of following：Welding current；Weldingvoltage；Speed of welding, the welding
Speed is defined as travel rate of the end of welding wire relative to soldered part on the direction of the weld seam；The welding wire
Feed rate；From contact nozzle to the distance of soldered part, the length of the distance instruction electric arc.
8. the welding system according to any one of claim 1 to 7, wherein, the welding data is to be welded except indicating
Also indicated outside material one or more of following：Welding technique to be used；Welding gas；Space for the weld seam
Physical dimension；The thickness of welding wire；Welding position.
9. the welding system according to any one of claim 1 to 8, wherein, the processing system is configured to be based on
The welding data selects adaptive algorithm among one group of adaptive algorithm.
10. a kind of method for training welding system, the welding system includes：
- welding actuator, the welding actuator operate according to actuator control data,
- sensing system, the sensing system are used to produce the sensing relevant with the weld seam as caused by the welding actuator
Device data, and
- processing system, the processing system are used for true using adaptive algorithm based on the sensing data and welding data
The fixed actuator control data, the welding data indicate material to be welded and welding condition,
Characterized in that, methods described includes：
- in order to train welding process, control number via training data interface input (201) actuator of the welding system
At least a portion actuator control data in, and
- according to the welding data, via the training data interface input the actuator control data in described in extremely
Lack a part of actuator control data and the sensing data measured during the training welding process to train
(201) adaptive algorithm, train welding process relative with described to be trained to the adaptive algorithm to control
The welding process answered.
11. according to the method for claim 10, wherein, the sensing data indicates one or more of following：From
The weld seam and from the area measure beside the weld seam to temperature；The surface profile of caused weld seam.
12. the method according to claim 10 or 11, wherein, the adaptive algorithm employment artificial neural networks are realized,
The artificial neural network can control according to the welding data, via the actuator of training data interface input
At least a portion actuator control data in data and the biography measured during the training welding process
Sensor data learn.
13. the method according to any one of claim 10 to 12, wherein, the adaptive algorithm with fuzzy logic come
Realize, the actuator control that the fuzzy logic can input according to the welding data, via the training data interface
At least a portion actuator control data in data processed and described in being measured during the training welding process
Sensing data is adapted.
14. the method according to any one of claim 10 to 13, wherein, the actuator control data includes welding
Parameter, the welding parameter include one or more of following：Welding current；Weldingvoltage；Speed of welding, the welding speed
Degree is defined as travel rate of the end of welding wire relative to soldered part on the direction of the weld seam；The welding wire
Feed rate；From contact nozzle to the distance of soldered part, the length of the distance instruction electric arc.
15. the method according to any one of claim 10 to 14, wherein, the welding data treats wlding except instruction
Also indicated outside material one or more of following：Welding technique to be used；Welding gas；Space for the weld seam
Physical dimension；The thickness of welding wire；Welding position.
16. the method according to any one of claim 10 to 15, wherein, methods described includes：Based on the welding
Data select adaptive algorithm among one group of adaptive algorithm.
17. a kind of computer program for being used to train welding system, the welding system include：
- welding actuator, the welding actuator operate according to actuator control data,
- sensing system, the sensing system are used to produce the sensing relevant with the weld seam as caused by the welding actuator
Device data, and
- processing system is may be programmed, the programmable processing system is used to utilize based on the sensing data and welding data
Adaptive algorithm determines the actuator control data, and the welding data indicates material to be welded and welding condition.
Characterized in that, the computer program includes computer executable instructions, the computer executable instructions are used to control
The programmable processing system is made, so as to：
- in order to train welding process, received from the training data interface of the welding system in the actuator control data
At least a portion actuator control data, and
Described in-actuator the control data received according to the welding data, from the training data interface to
Lack a part of actuator control data and the sensing data measured during the training welding process to train
The adaptive algorithm, the weldering corresponding with the training welding process can be controlled so as to which the adaptive algorithm is trained to
Termination process.
18. a kind of computer program product, it is with good grounds that the computer program product includes coding
The non-volatile computer-readable medium of computer program described in claim 17.
Applications Claiming Priority (3)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| FI20155077A FI20155077A (en) | 2015-02-06 | 2015-02-06 | welding System |
| FI20155077 | 2015-02-06 | ||
| PCT/FI2016/050051 WO2016124818A1 (en) | 2015-02-06 | 2016-01-28 | Welding system with adaptive algorithm |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107427951A true CN107427951A (en) | 2017-12-01 |
Family
ID=55345843
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201680009100.8A Pending CN107427951A (en) | 2015-02-06 | 2016-01-28 | Utilize the welding system of adaptive algorithm |
Country Status (4)
| Country | Link |
|---|---|
| EP (1) | EP3254161A1 (en) |
| CN (1) | CN107427951A (en) |
| FI (1) | FI20155077A (en) |
| WO (1) | WO2016124818A1 (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108788560A (en) * | 2018-08-08 | 2018-11-13 | 经略智能科技（苏州）有限公司 | Welding system based on XGBoost machine learning models |
| CN108817724A (en) * | 2018-08-08 | 2018-11-16 | 经略智能科技（苏州）有限公司 | Welding method based on XGBoost machine learning model |
| CN110936077A (en) * | 2019-12-31 | 2020-03-31 | 南京衍构科技有限公司 | Method for generating surfacing path of membrane type water-cooled wall |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP6457473B2 (en) * | 2016-12-16 | 2019-01-23 | ファナック株式会社 | Machine learning apparatus, robot system, and machine learning method for learning operation of robot and laser scanner |
| US10875125B2 (en) | 2017-06-20 | 2020-12-29 | Lincoln Global, Inc. | Machine learning for weldment classification and correlation |
| CN112894080A (en) * | 2019-11-19 | 2021-06-04 | 中国石油天然气集团有限公司 | Method and device for controlling length of welding arc |
| EP3855261B1 (en) * | 2020-01-27 | 2024-05-15 | ABB Schweiz AG | Determining control parameters for an industrial automation device |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO1995007165A1 (en) * | 1993-09-11 | 1995-03-16 | Robert Bosch Gmbh | Resistance-welding method |
| US6636776B1 (en) * | 2001-07-09 | 2003-10-21 | Lincoln Global, Inc. | System and method for managing welding procedures and welding resources |
| CN1634676A (en) * | 2005-01-13 | 2005-07-06 | 上海交通大学 | Full-digital speed change control method for wire feeding for submerged arc welding |
| CN1640603A (en) * | 2004-01-17 | 2005-07-20 | 上海威特力焊接设备制造有限公司 | Welding machine output characteristic control method |
| CN1971459A (en) * | 2006-11-29 | 2007-05-30 | 河北科技大学 | Comprehensive automatic control system for welding process of high-frequency welded pipe |
| CN101021718A (en) * | 2006-11-14 | 2007-08-22 | 河北科技大学 | Double-layer roll welding pipe welding process control system and control method thereof |
| CN202106129U (en) * | 2011-03-17 | 2012-01-11 | 安徽理工大学 | Automatic track-follower for welding joints of bundle tube production line |
| CN104070253A (en) * | 2013-03-29 | 2014-10-01 | 发纳科美国公司 | Adaptive control of robotic laser braze/weld operation |
-
2015
- 2015-02-06 FI FI20155077A patent/FI20155077A/en not_active IP Right Cessation
-
2016
- 2016-01-28 EP EP16703822.3A patent/EP3254161A1/en not_active Withdrawn
- 2016-01-28 CN CN201680009100.8A patent/CN107427951A/en active Pending
- 2016-01-28 WO PCT/FI2016/050051 patent/WO2016124818A1/en active Application Filing
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO1995007165A1 (en) * | 1993-09-11 | 1995-03-16 | Robert Bosch Gmbh | Resistance-welding method |
| US6636776B1 (en) * | 2001-07-09 | 2003-10-21 | Lincoln Global, Inc. | System and method for managing welding procedures and welding resources |
| CN1640603A (en) * | 2004-01-17 | 2005-07-20 | 上海威特力焊接设备制造有限公司 | Welding machine output characteristic control method |
| CN1634676A (en) * | 2005-01-13 | 2005-07-06 | 上海交通大学 | Full-digital speed change control method for wire feeding for submerged arc welding |
| CN101021718A (en) * | 2006-11-14 | 2007-08-22 | 河北科技大学 | Double-layer roll welding pipe welding process control system and control method thereof |
| CN1971459A (en) * | 2006-11-29 | 2007-05-30 | 河北科技大学 | Comprehensive automatic control system for welding process of high-frequency welded pipe |
| CN202106129U (en) * | 2011-03-17 | 2012-01-11 | 安徽理工大学 | Automatic track-follower for welding joints of bundle tube production line |
| CN104070253A (en) * | 2013-03-29 | 2014-10-01 | 发纳科美国公司 | Adaptive control of robotic laser braze/weld operation |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108788560A (en) * | 2018-08-08 | 2018-11-13 | 经略智能科技（苏州）有限公司 | Welding system based on XGBoost machine learning models |
| CN108817724A (en) * | 2018-08-08 | 2018-11-16 | 经略智能科技（苏州）有限公司 | Welding method based on XGBoost machine learning model |
| CN110936077A (en) * | 2019-12-31 | 2020-03-31 | 南京衍构科技有限公司 | Method for generating surfacing path of membrane type water-cooled wall |
| CN110936077B (en) * | 2019-12-31 | 2021-11-26 | 南京衍构科技有限公司 | Method for generating surfacing path of membrane type water-cooled wall |
Also Published As
| Publication number | Publication date |
|---|---|
| EP3254161A1 (en) | 2017-12-13 |
| FI20155077A (en) | 2016-08-07 |
| WO2016124818A1 (en) | 2016-08-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107427951A (en) | Utilize the welding system of adaptive algorithm | |
| CN111702306B (en) | Time-of-flight camera for welding machine vision | |
| US10363628B2 (en) | Automatic and semi-automatic welding systems and methods | |
| CN106457468B (en) | Use the hybrid laser welding system and method for Liang Ge robot | |
| US11536669B2 (en) | Detecting machining errors of a laser machining system using deep convolutional neural networks | |
| EP1495827A2 (en) | Welding wire positioning system and method | |
| US20220011726A1 (en) | Monitoring a laser machining process using deep folding neural networks | |
| CN108027911A (en) | Promote the system and method that welding is service software | |
| JP6920557B2 (en) | A method for automatically determining the optimum welding parameters for performing work piece welding | |
| US20200406392A1 (en) | Evaluation apparatus, evaluation method, evaluation system, and non-transitory computer-readable storage medium | |
| WO2020251038A1 (en) | Repair weld control device and repair weld control method | |
| CN112828496B (en) | Real-time weld quality analysis system and method | |
| CN114025904A (en) | Repair welding inspection device and repair welding inspection method | |
| CN113302017A (en) | Method for detecting welding defects in arc welding and arc welding system | |
| US20230234153A1 (en) | Method for defining welding parameters for a welding process on a workpiece and welding device for carrying out a welding process on a workpiece with defined welding parameters | |
| KR20100011585A (en) | Apparatus for tracking welding line | |
| CN109926703A (en) | Welding position detection device, welding position detection method and welding robot system | |
| WO2022019013A1 (en) | Machine learning device, laminate molding system, machine learning method for welding condition, welding condition determination method, and program | |
| JP7369982B2 (en) | Repair welding equipment and repair welding method | |
| JP5461070B2 (en) | Abnormality detection method for laser welding system | |
| Gyasi et al. | Digitalized automated welding systems for weld quality predictions and reliability | |
| JP2023095820A (en) | Generation method of learning model, control device, learning model, welding system and program | |
| CN116940433A (en) | Welding quality inspection system | |
| Tukahirwa et al. | Influence of Process Parameters in Gas-Metal Arc Welding (GMAW) of Carbon Steels | |
| CN107705312A (en) | A kind of method based on line scan data extraction postwelding weld edge point |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20171201 |
|
| WD01 | Invention patent application deemed withdrawn after publication |