Welding system and method for operating a welding system
Through the detection and evaluation device in the welding system, the observer model is used to monitor the welding process signals in real time and adjust the control parameters dynamically, the weld quality problems caused by irregular geometric shapes of the components are solved, and the efficient stability of the welding process and the welding gun are achieved. Accessibility.
Patent Information
- Application Number
- CN202210171429.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Priority Date
- 2021-02-25
- Filing Date
- 2022-02-24
- Publication Date
- 2024-11-12
- Estimated Expiration
- 2042-02-24
AI Technical Summary
During the welding process, due to manufacturing tolerances, the geometric shape irregularities of the component are difficult to adjust the control parameters in real time, resulting in reduced weld quality and defects, increasing production costs and rework risks.
Using the detection device and evaluation device in the welding system, the welding process signal is monitored in real time through the observer model, the characteristic values of the welding torch position and gap are determined, and the control parameters are dynamically adjusted according to these values, avoiding the smoothing filter or integrator caused by Delay, use current sensors, voltage sensors, solid acoustic sensors, etc. to detect welding process signals, and combine them with Longberg observer or Kalman filter models to simulate noise to determine characteristic values.
Real-time dynamic adjustment of the welding process is achieved, weld quality is improved, scrap and rework is reduced, manufacturing tolerances for component geometry is expanded, additional sensors and measurement steps are avoided, and the accessibility of the welding gun is maintained.
Smart Images

Figure CN114951906B_ABST
Abstract
Description
Technical Field
[0001] The invention relates to a welding system for carrying out a welding process in which at least two components are joined by a weld seam. The invention also relates to a method for operating such a welding system. Background Art
[0002] Gas metal arc welding is a very complex process due to the many joining parameters and influencing factors. The positioning of the welding torch / burner on the metal sheets to be welded (torch position / torch offset) and the gap width between the joining partners have a significant influence on the welding process. Depending on the relative position of the welding torch to the workpiece and the gap width, the optimal control parameters for the welding process can be determined experimentally. It should be noted that the term "welding torch" is also used here to refer to electric welding.
[0003] However, due to manufacturing tolerances, the workpiece geometry as well as the actual positioning of the welding gun relative to the parts to be welded and the actual gap width may fluctuate in series production. In the case of static control parameters, deviations from the ideal values or exceeding of limit values can lead to reduced weld quality or even defects in the weld joint. Defective weld joints mean scrap or additional rework, both of which are accompanied by additional costs. Usually, the control parameters of the welding process equipment are configured once at commissioning and the start of series production. This is regarded as the determination of the ideal / initial parameters. Here, an attempt is made to configure the tolerance range of the control parameters to the largest possible tolerance range with regard to the welding gun positioning and the gap bridging capability. However, not all manufacturing tolerances can be depicted, so the occurrence of weld irregularities cannot be completely avoided. This in turn requires downstream quality monitoring and the possibility of reworking the joint.
[0004] The dynamic adjustment of the control parameters presupposes knowledge of the individual geometry of the individual components, which requires measuring the components prior to the welding process. In order to detect irregularities in the weld seam during the manufacturing process according to the current prior art, considerable expenditure is required. For example, upstream optical detection by a camera and subsequent image processing can be carried out for this purpose. This approach requires an optical camera system or a laser triangulation sensor to accurately measure the individual component geometry of each weld seam. This approach has four major disadvantages. First, additional sensors are required, which incurs costs. Second, additional upstream steps for measuring the components are required, which increases production time. Third, an additional robot is required to perform the measurement of the component, or fourth, the sensor system must be mounted directly on the welding gun, which reduces the robot's accessibility to the weld seam, so that many typical welds of body structures can no longer be welded or the range of application is limited.
[0005] From US Pat. No. 4,943,701 A and US Pat. No. 4,435,631 A, it is known to observe the positioning of the welding torch relative to the workpiece during the welding process based on process signals and to provide feedback for dynamically adjusting control parameters. In this solution, a microphone detects the sound level of the airborne sound caused by the arc. The arc length derived from this sound level is a measure of the distance between the welding torch tip and the welding point. Since the microphone recordings are very noisy, the microphone signal must be smoothed using a smoothing filter. The signal delay caused by the smoothing filter can have a negative impact on the rapidity of the response of the control loop.
[0006] From document DE 10 2018 009 536 A1 it is known to estimate the current position of a handheld welding torch based on structure-borne sound signals.
[0007] From document DE 698 16 839 T2 it is known to estimate the gap width during the welding process based on the welding voltage which the welding process device requires to generate the welding current. Here too, an integrator must be used to smooth the measurement signal, so that the delay described in tracking the manipulated variable occurs in the process control. Summary of the invention
[0008] The object of the invention is to respond to irregularities in the geometry of the components during the welding process, if these irregularities are caused, possibly by manufacturing tolerances.
[0009] One aspect of the invention relates to a welding system. The welding system comprises a welding process device for performing at least one welding process under consideration of predetermined control parameters, wherein a weld seam is produced between at least two parts to be joined or connected in the corresponding welding process, and the so-called welding torch / burner of the welding process device has a relative welding torch position relative to the parts and / or relative to the gap formed between the parts. The welding process device can be designed for gas shielded metal arc welding (MSG). In particular, an automatic welding machine is provided as the welding process device, wherein the welding torch and / or at least one part is guided automatically, robotically or manually. In order to achieve the predetermined control parameters, the welding process device can have a regulator. For example, the current intensity and / or voltage for the arc or welding wire, the welding torch speed along the welding line, the wire feeding speed of the welding wire and / or the welding torch position can be predetermined as control parameters.
[0010] In order to monitor the welding process, the welding process device also includes a detection device for detecting a current process signal occurring during the welding process and an evaluation device, the detection device being configured to determine from the current process signal at least one corresponding current characteristic value of the current torch position and / or the gap existing between the components. An example of a characteristic value of the torch position is a distance of the torch from at least one of the components resulting from the torch position. The process signal can be, for example, a measured current intensity or a measured voltage or, in general, a measurement signal detected by a measuring circuit. The time profile of a control parameter can also be used as a process signal.
[0011] As described at the outset in conjunction with the prior art, it is provided that the welding process is monitored in such a way that a process signal is detected which depends on the current state and / or the current progress of the welding process or is characteristic of the welding process. However, as already mentioned above, in order to determine at least one characteristic value from such a process signal, the prior art provides for smoothing or integrating the process signal with the aid of a smoothing filter or an integrator, which, however, results in a delay or, in a control-technical manner, results in a numerically large phase. The prior art does not use a model for the relationship between the process signal and the characteristic value. Therefore, the characteristic value must be explicitly detected with the aid of an additional measuring technique (e.g. a camera) or extracted with the aid of an upstream or downstream process, for example by forming a mean value.
[0012] In order to avoid this, the evaluation device is provided according to the invention for determining at least one characteristic value by running an observer model of the welding process and determining the at least one characteristic value based on the process signal with the aid of the observer model. Such an observer model can be a component of a control observer, such as a so-called Luenberger observer or a Kalman filter. By simulating the welding process by the observer model, noise characteristic of the welding process can also be simulated, so that, for example, smoothing filters or integrators are not required, since the noise is part of the modeling. The at least one characteristic value can be updated as a function of the process signal without the associated smoothing constant causing a delay.
[0013] The invention also includes embodiments that produce additional advantages.
[0014] One embodiment includes: the detection device for detecting the current process signal generated during the welding process comprises at least a current sensor and / or a voltage sensor and / or an airborne sound microphone and / or a structure-borne sound sensor and / or an ultrasonic sensor. The current intensity of the arc current or the current in the welding wire can be detected by the current sensor. This is a corresponding measure of the welding torch spacing. In addition or alternatively, the gap width of the gap can be inferred. The voltage sensor can be used to generate and / or maintain the melting process at the weld seam. The voltage required for this purpose can be detected by the voltage sensor. The noise caused by the welding process, such as the arc noise, can be detected by the airborne sound microphone. Its loudness level depends on the length of the arc and thus also on the distance between the welding torch and the current welding point. The structure-borne sound and / or vibration caused by the welding process in at least one component and / or the welding torch can be detected by the structure-borne sound sensor. The loudness level and / or phase shift between the sensor signals of two or more structure-borne sound sensors can also be used to determine at least one characteristic value, in particular the distance from the welding torch to the gap and / or the gap width. Compared to the above-described optical camera systems, airborne sound microphones and / or structure-borne sound sensors and / or ultrasonic sensors have the advantage that they are significantly smaller in design and do not have to be positioned on the welding torch, thereby not restricting accessibility.
[0015] One embodiment includes providing a mechanical clamping device for at least one of the components and arranging at least one structure-borne sound sensor of the detection device on the clamping device. If the structure-borne sound is detected indirectly by means of a mechanical clamping device, this has the advantage that the structure-borne sound sensor can always remain mounted on the clamping device and thus does not need to be provided when at least one component is clamped for a welding process. It is also possible to arrange at least one airborne sound microphone and / or at least one ultrasonic sensor on the clamping device.
[0016] According to one embodiment, a process signal is measured and an observer model calculates a characteristic value or at least one characteristic value predicted according to the model from the process signal. These actual characteristic values are compared with target characteristic values / theoretical characteristic values, for example from a process plan or a design plan, and if a deviation is detected, a control parameter is adjusted to reduce the deviation or to adapt the welding process to the actual characteristic value. In other words, it is provided that the evaluation device is configured to calculate at least one characteristic value from the measured process signal by means of the observer model, compare the at least one characteristic value with the corresponding target characteristic value, and if a deviation is detected, adjust the control parameter to reduce the deviation or to adapt the welding process to the at least one characteristic value.
[0017] One embodiment comprises that in the observer model at least one characteristic value is set as a parameter of the observer model, and the observer model estimates at least one of the process signals by means of the observer model, and an estimated deviation between the estimated and detected process signal is determined from a comparison of the respective estimated process signal with the corresponding actually detected process signal, which estimated deviation is minimized by adjusting the respective characteristic value in the observer model. Thus, it is predicted or estimated by means of the observer model which time distribution of the at least one predicted process signal is likely to occur when the at least one characteristic value is correctly set in the observer model. If this is not the case, it can be determined, for example by iteratively varying the at least one characteristic value in the observer model, for which said at least one characteristic value the estimated deviation is minimized, which is then used as the optimal estimate for the at least one characteristic value.
[0018] In addition or as an alternative to this, according to one embodiment, the observer model is designed as an artificial neural network. Thus, in the observer model, a mapping function for the signal value of at least one process signal onto at least one characteristic value can be created by means of the artificial neural network. For this purpose, training data can be generated, which also indicate known characteristic values relative to known process signals, such as can be determined, for example, in a prototype or test laboratory by means of a sensor system. If the artificial neural network is trained by a machine learning method using these experimentally determined and therefore known correlations between process signals and characteristic values, characteristic values that are still unknown can then be determined by means of the artificial neural network, in particular in real time, based on the process signals detected during the welding process.
[0019] It can also be additionally provided for the observer model that the regulator of the welding process device is set to feed at least one regulator signal as an input signal into the observer model. In addition or as an alternative to this, a manipulated variable signal of at least one manipulated variable of the welding process device, for example a currently set wire feed speed, can be fed into the observer model as an input signal. Since the observer model simulates the welding process, the reaction of the welding process to at least one regulator signal and / or at least one manipulated variable signal can be determined based on at least one regulator signal and / or at least one manipulated variable signal by means of the observer model, so that the process signal or at least one of the process signals is predicted or estimated by the observer model as an output. The prediction and / or estimation depends on at least one characteristic value implemented as a parameter in the observer model. By determining the deviation of the estimated at least one process signal from the corresponding actually detected process signal, the optimization or adaptation of at least one characteristic value can be performed in the described manner.
[0020] So far it has been described how at least one characteristic value of the torch position and / or gap can be determined from the process signal based on an observer model. In the following, a possible use of the estimated at least one characteristic value, which is enabled by embodiments of the present invention, is described.
[0021] The at least one characteristic value can be used in different ways. It can be provided that the welding process itself is controlled thereby and / or at least another manufacturing process or manufacturing step can be controlled and / or quality monitoring can be performed. All of this will be explained in more detail below.
[0022] One embodiment comprises: providing a regulating device for regulating a control parameter and the regulating device is arranged to determine the control parameter based on at least one characteristic value and based on a characteristic curve of at least one control parameter provided in the regulating device and to predetermine the control parameter at the welding process as a target value for a regulator of the welding process device. In other words, a higher-level regulating loop or control loop is provided, which determines a target value for an internal regulator of the welding process device that is suitable for the welding process device so that the current actual geometric conditions of at least one component can be taken into account. In this way, the corresponding target value of the regulator of the welding process device can be updated continuously or cyclically with a predetermined sampling time. Which target value is optimal or advantageous for which of the at least one characteristic value can be determined by means of a characteristic curve known per se. The characteristic value can be mapped to the corresponding value of the control parameter by means of the characteristic curve. Suitable characteristic curves are available in the prior art, which, for example, map the gap width to the welding current. In addition, at least one component information related to at least one component to be welded can be taken into account, which can be determined, for example, based on CAD data (CAD-computer-aided design) or generally by design data for the component.
[0023] One embodiment includes: the control parameter includes at least one of the following parameters: target torch position, target current intensity, target voltage, target torch speed, target welding speed, target wire feeding speed. Thus, the torch position itself can be adjusted, but it is also possible, for example, not to track or adjust the torch position according to at least one characteristic value, but to adapt the current intensity and / or voltage and / or torch speed and / or wire feeding speed to a given torch position, which advantageously achieves a reduction in response time. However, it is also possible to predetermine at least one of the torch position and other given variables as control parameters.
[0024] However, at least one characteristic value can also be used outside the actual welding process. One embodiment comprises providing at least one manufacturing device, which is configured to perform a manufacturing step different from the welding process for at least one of the components, and the manufacturing device and / or the evaluation device is configured to dynamically configure the corresponding manufacturing step during operation according to the at least one characteristic value. Thus, at least one characteristic value can be used outside the welding process in at least one other manufacturing step in order to dynamically adapt it to or track the at least one characteristic value determined or observed during operation. The corresponding manufacturing step can be arranged upstream or downstream of the welding process. For example, as an upstream manufacturing step, at least one component can be cut, and when, for example, a deviation of the geometry of the cut component from the target geometry is identified according to the at least one characteristic value, this cutting can be adjusted or configured. This deviation is reduced by the adjustment.
[0025] One embodiment comprises: the recording device is arranged to record the detected deviation of at least one characteristic value from the corresponding target characteristic value based on at least one characteristic value for quality monitoring of the welding process and / or quality monitoring of at least one manufacturing process upstream of the welding process. In other words, the quality of the welding process and / or the quality of at least one other manufacturing process upstream of the welding process can also be recorded based on the at least one characteristic value. Thus, complex, for example manual quality control can be replaced by determining at least one characteristic value. The variable associated with the corresponding characteristic value can be, for example, a variable that is proportional to the corresponding characteristic value (for example torch spacing and / or gap width) or is linked by another previously known function.
[0026] One embodiment comprises: at least one characteristic value indicates or depends on the distance between the welding torch and at least one component and / or the distance between the welding torch and the gap and / or the distance between the welding torch and the molten material and / or the gap width of the gap. It has been shown that the characteristic value can be determined based on process signals, in particular current intensity and / or voltage and / or structure-borne sound emission. The observer model required for this purpose can be formed based on physical relationships known per se and / or by training an artificial neural network and / or based on characteristic curves that can be predetermined in tests.
[0027] One aspect of the invention relates to a method for operating a welding system, wherein a welding process device performs at least one welding process taking into account predetermined control parameters, wherein in the respective welding process a weld seam is produced between at least two components to be connected and at the same time a welding torch of the welding process device has a relative welding torch position relative to the components and / or relative to a gap formed between the components. A detection device detects a current process signal occurring during the welding process and an evaluation device determines from the current process signal at least one corresponding current characteristic value of the current welding torch position and / or gap. To determine at least one characteristic value, the evaluation device runs an observer model of the welding process and determines at least one characteristic value based on the process signal with the aid of the observer model. The method thus results from the operation of the described welding system. The method has the advantage that at least one characteristic variable can be determined in real time or online during the welding process and that the welding process can thus be observed and / or the welding process can be adapted to the respective at least one current characteristic value by feedback or adjustment of control parameters.
[0028] The invention also includes improvements of the method according to the invention which have the features already described in conjunction with the improvements of the welding system according to the invention. For this reason, corresponding improvements of the method according to the invention are not described here.
[0029] The evaluation device can be provided on the basis of a data processing device or a processor device, which is configured to perform the steps described in conjunction with the evaluation device. For this purpose, the processor device can have at least one microprocessor and / or at least one microcontroller and / or at least one FPGA (field programmable gate array) and / or at least one DSP (digital signal processor). In addition, the processor device can have a program code, which is configured to perform the method steps when executed by the processor device. The program code can be stored in a data memory of the processor device. The described adjustment device can be configured accordingly.
[0030] The invention also includes combinations of features of the described embodiments. As long as these embodiments are not described as mutually exclusive, the invention also includes embodiments each having a combination of features of a plurality of the described embodiments. BRIEF DESCRIPTION OF THE DRAWINGS
[0031] The following describes an embodiment of the present invention. The only accompanying drawing shows:
[0032] Figure 1 A schematic diagram showing an embodiment of a welding system according to the present invention, which welding system can perform an embodiment of the method according to the present invention; and
[0033] Figure 2A schematic diagram is shown for illustrating characteristic variables of a welding process.
[0034] The embodiments explained below are preferred embodiments of the present invention. In the embodiments, the described components of the embodiments are respectively single features of the present invention that can be regarded as independent of each other, and these features further improve the present invention independently of each other. Therefore, the disclosure should also include feature combinations other than the feature combinations of the embodiments shown. In addition, the described embodiments can also be supplemented by other features of the present invention that have been described.
[0035] In the figures, the same reference numerals respectively denote elements with the same function. DETAILED DESCRIPTION
[0036] Figure 1 A welding system 10 is shown, by which a welding process 11 can be performed, for example, when producing a workpiece. In this case, the welding process 11 can be performed automatically by a welding process device 12. Figure 2 1 shows the welding process 11 in more detail. The welding process 11 can provide, for example, for producing a weld seam 16 for joining or connecting the components 14 between at least two joining partners or components 14 in a gap 15 defined by them by means of a gas-shielded metal arc welding by means of an electric current contact nozzle or welding torch 13. For this purpose, in a manner known per se, the welding torch 13 can melt a welding wire 17 in the region of a melting point or in the region of a molten material 18 by means of an electric arc and / or by means of an electric current or liquefy together with the material of the components 14 in the gap 15. For this purpose, the welding torch 13 can have a welding torch offset or a spacing 19 from the gap 15. The gap 15 can have a gap height or generally a gap width 20. The gap width 20 can deviate from a planned or prescribed gap width, since the production of the components 14 can have tolerances, so that, for example, the shape or contour of the edges of the components 14 can be affected by fluctuations or deviations.
[0037] like Figure 1 As further shown in FIG. 1 , the components 14 may be provided, for example, by an upstream manufacturing device 21 , which may, for example, cut at least one of the components 14 , thereby causing fluctuations or tolerances. The manufacturing device 21 may, for example, be a device for cutting or stamping at least one component 14 .
[0038] The welding process device 12 can have, for example, a regulator 22, by means of which a manipulated variable signal or a manipulated variable parameter 23 can be set for controlling or regulating the welding process 11. For example, this can be a torch speed and / or a wire feed speed and / or a torch positioning and / or a current intensity and / or a voltage along the gap 15 for the welding process 11. In the welding process 11, a manipulated variable 24 can be provided in a manner known per se in a feedback loop 25 for the regulator 22 by at least one sensor circuit. One or several or all of the manipulated variables 24 can be provided as a process signal 26 to an evaluation device 27. The evaluation device 27 can be based on a processor circuit, for example.
[0039] An observer 28 with a process model or observer model 29 of the welding process 11 can be operated by the evaluation device 27 in terms of control technology. For example, a Romberg observer can be provided as the observer 28. The observer model 29 can describe the welding process 11, for example, based on a system of equations and / or based on a characteristic curve of the welding process 11 and / or based on an artificial neural network. A suitable observer model 29 can be determined in a manner known per se based on a test series of the welding process 11. The evaluation device 27 can estimate or determine at least one characteristic value 30 of the welding process 11 with the aid of the observer model 29. For example, the gap width 20 and / or the spacing 19 can be determined as characteristic values.
[0040] In addition to the process signal 26, at least part of the manipulated variable parameter 23 can also be provided as an input signal to the observer model 29. In addition or as an alternative thereto, component information 33, ie component data, can also be provided to the observer model 29 as an input signal.
[0041] Additionally or alternatively, the observer model 29 may include an artificial neural network which maps the process signal 26 as an input variable onto at least one estimated characteristic value 30 as an output variable.
[0042] For example, at least one characteristic value 30 can be provided to the manufacturing device 21 in order to configure a control device and / or a regulating device 31 on the manufacturing device in order to, for example, match manufacturing errors or deviations in the geometry of the component 14 provided by the manufacturing device 21 to a corresponding target geometry, so that a future component 14 with a corrected geometry is provided by the manufacturing device 21.
[0043] The at least one characteristic value 30 can additionally or alternatively be provided to a regulating device 32, which can also be operated, for example, on a processor circuit or on the same processor circuit as the evaluation device 27. The regulating device 32 can provide an updated value of a control parameter 34 for the welding process device 12, for example based on component information 33 related to the at least two components 14 to be welded or joined and based on the at least one characteristic value 30. The control parameter 34 can serve as a target value for the regulator 22. Thus, by setting the control parameter 34, the regulator 22 can compensate for component tolerances or changes in the component 14 by setting or tracking the at least one manipulated variable parameter 23.
[0044] The control of the controller 22 of the welding process device 12 during the welding process 11 therefore represents the inner control loop 35 of the process control, while the at least one process signal 26 fed back into the control device 32 via the evaluation device 27 represents the outer control loop.
[0045] Thus, this concept includes a device / arrangement of a welding system 10 with a plurality of sensors that detect process data or process signals 26 during the welding process. The relative positioning of the welding torch relative to the workpiece (together with its components 14) is reconstructed from the process data or process signals in particular. The process data or process signals 26 preferably include the reception / recording of current and voltage values from the welding system and / or the reception of sound emission values (in particular structure-borne sound). Airborne sound and / or ultrasound can also be used in addition or as an alternative. The observed torch positioning (torch position with a spacing 19) is then fed back as a control variable into a superimposed control loop during the welding process for adjusting the control parameters 34. In this context, it is particularly concerned with expanding the classical process control by further higher-level control loops. Thus, the device / arrangement enables dynamic adjustment of the control parameters 34 during the welding process 11 without causing additional production time and without reducing the accessibility of the welding gun and its welding torch 13 due to additional optical sensors on the robot guiding the welding torch 13.
[0046] The device of the welding system 10 preferably includes at least the following components:
[0047] 1. Welding process equipment;
[0048] 2. Welding process;
[0049] 3. Sensors for detecting the welding process (current, voltage, structure-borne sound, etc. are only examples);
[0050] 4. Observer with observer model;
[0051] 5. Regulators for controlling parameters;
[0052] The advantages of the welding system are:
[0053] Dynamically adjust control parameters 34 during the welding process 11;
[0054] Additional equipment on the welding gun (or welding torch 13 for short) does not restrict accessibility;
[0055] · Allowing for higher manufacturing tolerances with respect to the component geometry of the component 14;
[0056] Reduce scrap and rework;
[0057] The detected deviations of the welding torch position (particularly the distance 19 ) and the workpiece and its component 14 can be used for quality monitoring of the manufacturing process upstream of the manufacturing device 21 .
[0058] The device preferably comprises the above-mentioned components 1 to 5. The figures show a schematic arrangement.
[0059] The controller of the regulating device 32 for the control parameter 34 contains in particular a characteristic curve for an optimum value of the control parameter 34 based on the component information 33 and the torch position, which can be provided as a characteristic value 30, for example as a torch offset or spacing 19. Based on the observed torch position, the controller continuously determines new control parameters 34 for the welding process 11. The control parameters 34 are transmitted to the welding process device 12 in real time as target values.
[0060] The welding process device 12 regulates the welding process 11 . The regulating behavior or regulating characteristic of the regulator 22 depends on the control parameter 34 and a characteristic curve, which can be adjusted during the welding process 11 , for example based on the control parameter 34 .
[0061] The welding process 11 is detected by recording 1 to n process signals 26 (n being the total number of detected process signals 26). The detected process signals 26 are in particular the current and voltage of the arc and signals from other external sensors which do not have to be attached to the welding gun, in particular to the welding torch 13. For example, a suitable sensor may be a structure-borne sound sensor 36 which is attached to a clamping device 37 for the respective component 14. The clamping device 37 is located at Figure 2 The process signal 26 is transmitted in real time to an observer model 29 of an observer 28 in control technology.
[0062] The observer 28 contains an observer model 29 of the welding process 11. The observer 28 calculates the relative torch position, for example the distance 19, relative to the workpiece and its components 14 during the welding process 11 from the process signal 26. The observed torch position is then transmitted in real time to a controller of a control device 32 for controlling parameters 34.
[0063] A preferred corresponding embodiment for the feedback of the welding torch position as a manipulated variable during a welding process is characterized in particular by a welding system comprising:
[0064] - a welding process device for carrying out a welding process according to respectively defined process parameters taking into account predetermined control parameters,
[0065] wherein in a corresponding welding process a weld seam is produced between two components 14 to be connected, which are separated by a corresponding gap with a corresponding gap width / gap height,
[0066] - means for detecting current process signals generated during the welding process,
[0067] - a regulating device for regulating a predetermined process parameter as a function of the detected current process signal,
[0068] It is characterized in that an evaluation device is provided, which is configured to determine at least one corresponding current characteristic value for each current torch position / current torch offset and / or gap width from a current process signal; and a regulation device is provided, which is configured to adjust or adapt a given control parameter also while taking into account the corresponding current characteristic value.
[0069] The welding system is characterized in particular in that the device for detecting current process signals occurring during the welding process comprises at least one current sensor, one voltage sensor and / or one structure-borne sound sensor.
[0070] In general, these examples show how an apparatus / arrangement may be provided for feeding back torch position and / or gap width as controlled variables into the welding process.
Claims
1. A welding system (10), wherein include: - a welding process device (12) for carrying out at least one welding process (11) taking into account predetermined control parameters (34), -in, In a corresponding welding process (11), a weld seam (16) is produced between at least two components (14) to be connected, a welding torch (13) of a welding process device (12) having a relative welding torch position with respect to the components (14) and / or with respect to a gap (15) formed between the components (14), - a detection device for detecting a current process signal (26) occurring during the welding process (11), an evaluation device (27) which is configured to determine at least one corresponding current characteristic value (30) of a current welding torch position and / or gap (15) from a current process signal (26), It is characterized in that at least one characteristic value (30) is indicative of or depends on a distance (19) between the welding torch (13) and at least one component (14) and / or a distance between the welding torch and the gap (15) and / or a distance between the welding torch and the molten material (18) and / or a gap width (20) of the gap (15), The evaluation device (27) is configured to run an observer model (29) of the welding process (11) in order to determine at least one characteristic value (30) and to determine at least one characteristic value (30) based on the process signal (26) by means of the observer model (29). The observer model (29) describes the relationship between the process signal and the characteristic value, In an observer model (29) for calculating at least one characteristic value (30), the at least one characteristic value is set as a parameter of the observer model (29), the observer model (29) estimates at least one process signal (26) at the estimated characteristic value, and based on a comparison of the respective estimated process signal (26) with the corresponding detected process signal (26), determines an estimated deviation between the estimated and detected process signals (26), and minimizes this estimated deviation by adjusting the respective characteristic value (30).
2. The welding system (10) according to claim 1, It is characterized in that The observer model (29) is designed as an artificial neural network.
3. The welding system (10) according to claim 1 or 2, It is characterized in that A detection device for detecting a current process signal (26) occurring during a welding process (11) comprises at least a current sensor and / or a voltage sensor and / or an airborne sound microphone and / or a structure-borne sound sensor (36) and / or an ultrasonic sensor.
4. The welding system (10) according to claim 3, It is characterized in that A mechanical holding device (37) is provided for at least one of the components (14), and at least one structure-borne sound sensor (36) and / or at least one airborne sound microphone and / or at least one ultrasonic sensor of the detection device are arranged on the holding device (37).
5. The welding system (10) according to claim 1 or 2, It is characterized in that The evaluation device (27) is configured to calculate at least one characteristic value (30) from the measured process signal with the aid of an observer model (29) and to compare the at least one characteristic value with a corresponding target characteristic value and, if a deviation is detected, to adjust a control parameter (34) in order to reduce the deviation and / or to adapt the welding process (11) to at least one current characteristic value.
6. The welding system (10) according to claim 1 or 2, It is characterized in that A regulating device (32) for regulating a control parameter (34) is provided, the regulating device being configured to determine the control parameter (34) as a function of at least one characteristic value (30) and as a function of at least one characteristic curve for the control parameter (34) provided in the regulating device (32), and to pre-set the control parameter for a welding process device (12) as a target value for a regulator (22) of the welding process device (12).
7. The welding system (10) according to claim 1 or 2, It is characterized in that The control parameters (34) include at least one of the following parameters: target torch position, target current intensity, target voltage, target torch speed, target welding speed, and target wire feeding speed.
8. The welding system (10) according to claim 1 or 2, It is characterized in that At least one manufacturing device (21) is provided, which is configured to perform a manufacturing step different from a welding process (11) for at least one component (14), the manufacturing device (21) and / or an evaluation device (27) being configured to dynamically configure the corresponding manufacturing step during operation based on at least one characteristic value (30).
9. The welding system (10) according to claim 1 or 2, It is characterized in that The recording device is configured to record, based on at least one characteristic value (30), a detected deviation of the at least one characteristic value (30) from a corresponding target value for quality monitoring of a welding process (11) and / or quality monitoring of at least one manufacturing process upstream of the welding process (11).
10. A method for operating a welding system (10), in, The welding process device (12) performs at least one welding process (11) taking into account predetermined control parameters (34), wherein in the respective welding process (11) a weld seam (16) is produced between at least two components (14) to be connected, the welding torch (13) of the welding process device (12) having a relative welding torch position relative to the components (14) and / or relative to a gap (15) formed between the components (14), The detection device detects a current process signal occurring during the welding process (11), The evaluation device (27) determines from the current process signal (26) at least one corresponding current characteristic value of the current welding torch position and / or the gap (15), The method is characterized in that at least one characteristic value (30) indicates or depends on a torch offset or a distance (19) between the welding torch (13) and at least one component (14) and / or a distance between the welding torch and a gap (15) and / or a distance between the welding torch and molten material (18) and / or a gap width (20) of the gap (15), and that in order to determine the at least one characteristic value (30), the evaluation device (27) runs an observer model (29) of the welding process (11) and determines the at least one characteristic value (30) based on the process signal (26) by means of the observer model (29), The observer model (29) describes the relationship between the process signal and the characteristic value, In an observer model (29) for calculating at least one characteristic value (30), the at least one characteristic value is set as a parameter of the observer model (29), the observer model (29) estimates at least one process signal (26) at the estimated characteristic value, and based on a comparison of the respective estimated process signal (26) with the corresponding detected process signal (26), determines an estimated deviation between the estimated and detected process signals (26), and minimizes this estimated deviation by adjusting the respective characteristic value (30).
11. The method according to claim 10, It is characterized in that The observer model (29) is designed as an artificial neural network.
