EP4122042A1 - Method and test facility for testing a bipolar plate of an electrochemical cell, in particular of a fuel cell - Google Patents

Method and test facility for testing a bipolar plate of an electrochemical cell, in particular of a fuel cell

Info

Publication number
EP4122042A1
EP4122042A1 EP21717314.5A EP21717314A EP4122042A1 EP 4122042 A1 EP4122042 A1 EP 4122042A1 EP 21717314 A EP21717314 A EP 21717314A EP 4122042 A1 EP4122042 A1 EP 4122042A1
Authority
EP
European Patent Office
Prior art keywords
plate
testing
test
bipolar plate
suspect
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
Application number
EP21717314.5A
Other languages
German (de)
French (fr)
Inventor
Christoph Hildebrandt
Chris CHOUTKA
Jan Achtzehn
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Schaeffler Technologies AG and Co KG
Original Assignee
Schaeffler Technologies AG and Co KG
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Schaeffler Technologies AG and Co KG filed Critical Schaeffler Technologies AG and Co KG
Publication of EP4122042A1 publication Critical patent/EP4122042A1/en
Pending legal-status Critical Current

Links

Classifications

    • H—ELECTRICITY
    • H01—ELECTRIC ELEMENTS
    • H01M—PROCESSES OR MEANS, e.g. BATTERIES, FOR THE DIRECT CONVERSION OF CHEMICAL ENERGY INTO ELECTRICAL ENERGY
    • H01M8/00—Fuel cells; Manufacture thereof
    • H01M8/02—Details
    • H01M8/0202—Collectors; Separators, e.g. bipolar separators; Interconnectors
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
    • G01N21/84—Systems specially adapted for particular applications
    • G01N21/88—Investigating the presence of flaws or contamination
    • G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
    • 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
    • B23K31/00—Processes relevant to this subclass, specially adapted for particular articles or purposes, but not covered by only one of the preceding main groups
    • B23K31/12—Processes relevant to this subclass, specially adapted for particular articles or purposes, but not covered by only one of the preceding main groups relating to investigating the properties, e.g. the weldability, of materials
    • B23K31/125—Weld quality monitoring
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
    • G01N21/84—Systems specially adapted for particular applications
    • G01N21/88—Investigating the presence of flaws or contamination
    • G01N21/95—Investigating the presence of flaws or contamination characterised by the material or shape of the object to be examined
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N23/00—Investigating or analysing materials by the use of wave or particle radiation, e.g. X-rays or neutrons, not covered by groups G01N3/00 – G01N17/00, G01N21/00 or G01N22/00
    • G01N23/02—Investigating or analysing materials by the use of wave or particle radiation, e.g. X-rays or neutrons, not covered by groups G01N3/00 – G01N17/00, G01N21/00 or G01N22/00 by transmitting the radiation through the material
    • G01N23/06—Investigating or analysing materials by the use of wave or particle radiation, e.g. X-rays or neutrons, not covered by groups G01N3/00 – G01N17/00, G01N21/00 or G01N22/00 by transmitting the radiation through the material and measuring the absorption
    • G01N23/18—Investigating the presence of flaws defects or foreign matter
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N27/00—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means
    • G01N27/72—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means by investigating magnetic variables
    • G01N27/82—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means by investigating magnetic variables for investigating the presence of flaws
    • G01N27/83—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means by investigating magnetic variables for investigating the presence of flaws by investigating stray magnetic fields
    • G01N27/84—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means by investigating magnetic variables for investigating the presence of flaws by investigating stray magnetic fields by applying magnetic powder or magnetic ink
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N29/00—Investigating or analysing materials by the use of ultrasonic, sonic or infrasonic waves; Visualisation of the interior of objects by transmitting ultrasonic or sonic waves through the object
    • G01N29/04—Analysing solids
    • G01N29/043—Analysing solids in the interior, e.g. by shear waves
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N29/00—Investigating or analysing materials by the use of ultrasonic, sonic or infrasonic waves; Visualisation of the interior of objects by transmitting ultrasonic or sonic waves through the object
    • G01N29/44—Processing the detected response signal, e.g. electronic circuits specially adapted therefor
    • G01N29/4481—Neural networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/0002—Inspection of images, e.g. flaw detection
    • G06T7/0004—Industrial image inspection
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
    • G01N21/84—Systems specially adapted for particular applications
    • G01N21/88—Investigating the presence of flaws or contamination
    • G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
    • G01N2021/8854—Grading and classifying of flaws
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
    • G01N21/84—Systems specially adapted for particular applications
    • G01N21/88—Investigating the presence of flaws or contamination
    • G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
    • G01N2021/8883—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges involving the calculation of gauges, generating models
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
    • G01N21/84—Systems specially adapted for particular applications
    • G01N21/88—Investigating the presence of flaws or contamination
    • G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
    • G01N2021/8887—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges based on image processing techniques
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
    • G01N21/84—Systems specially adapted for particular applications
    • G01N21/88—Investigating the presence of flaws or contamination
    • G01N21/91—Investigating the presence of flaws or contamination using penetration of dyes, e.g. fluorescent ink
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N2223/00—Investigating materials by wave or particle radiation
    • G01N2223/60—Specific applications or type of materials
    • G01N2223/629—Specific applications or type of materials welds, bonds, sealing compounds
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N2223/00—Investigating materials by wave or particle radiation
    • G01N2223/60—Specific applications or type of materials
    • G01N2223/646—Specific applications or type of materials flaws, defects
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N2291/00—Indexing codes associated with group G01N29/00
    • G01N2291/04—Wave modes and trajectories
    • G01N2291/044—Internal reflections (echoes), e.g. on walls or defects
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N2291/00—Indexing codes associated with group G01N29/00
    • G01N2291/26—Scanned objects
    • G01N2291/267—Welds
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N2291/00—Indexing codes associated with group G01N29/00
    • G01N2291/26—Scanned objects
    • G01N2291/269—Various geometry objects
    • G01N2291/2697—Wafer or (micro)electronic parts
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20081—Training; Learning
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20084—Artificial neural networks [ANN]
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
    • Y02E60/00—Enabling technologies; Technologies with a potential or indirect contribution to GHG emissions mitigation
    • Y02E60/30—Hydrogen technology
    • Y02E60/50—Fuel cells

Definitions

  • the invention relates to a method for testing a bipolar plate, in particular an electrochemical cell such as a fuel cell, and a test system provided for this purpose, with the aid of which a bipolar plate of an electrochemical cell such as a fuel cell can be examined for defects during manufacture.
  • a fuel cell stack or stack typically includes a plurality of fuel cells in a stack configuration.
  • the respective fuel cell comprises an electrolyte and electrodes, which are contacted via conductive plates.
  • a polymer electrolyte fuel cell that is operated in the low temperature range, a polymer membrane-electrode unit is present.
  • Electrically conductive bipolar plates mostly made of metal, are used to separate the individual polymer electrolyte fuel cells in a stack. They not only serve to make electrical contact with the electrodes and conduct the electrical current to the neighboring row, but also help to supply and distribute fuels and cooling media as well as to dissipate heat and reaction products.
  • bipolar plates usually have gas distribution fields which, due to their structure, lead to an optimal distribution of the gaseous fuels, mostly hydrogen and oxygen, with regard to the membrane surfaces.
  • the bipolar plate can be produced by two or more thin-walled metal sheets welded to one another, which form the desired flow paths, cooling channels and / or openings.
  • US 2019/0340747 A1 discloses a quality monitoring system and method in the area of a fuel cell production line.
  • US 2013/0230072 A1 describes a method and a device for error detection on fuel cell components, in particular bipolar plates.
  • DE 10393237 B4 discloses a method for detecting electrical defects in membrane electrode arrangements.
  • DE 3809221 A1 describes a method and a device for detecting defects, in particular cracks and / or constrictions, on pressed parts or other workpieces.
  • DE102015221 697 B3 discloses an arrangement for determining the surface quality of component surfaces, in particular of defects.
  • DE 102016211 449 A1 describes a test system with a portable test unit and a method for testing components.
  • DE 102009059765 A1 discloses a method for positioning a bipolar plate.
  • the bipolar plate is formed step by step from a first and a second plate, after each step an actual pattern of a large number of lines is generated by projection onto an intermediate product representing the respective step and recorded by means of a camera. It is then compared with a reference pattern.
  • One embodiment relates to a method for testing a bipolar plate of an electrochemical cell, in particular a fuel cell, in which an image of a surface of a bipolar plate is created, the image is examined for possible defects by an automated image processing-assisted evaluation system and in the event that the evaluation system an examined bipolar plate is identified as a potentially defective suspect plate, a detailed examination of an area of the suspect plate identified as potentially defective is carried out, the evaluation system displaying images of bipolar plates using a large number of non-defective and defective bipolar plates has a trained neural network for identifying a suspect plate and the neural network is continuously trained with the results of the detailed check, the neural network performing reinforcement learning.
  • a bipolar plate is understood to be a component of an electrochemical cell which provides the functions of an electrode and a flow guide element.
  • This can be, for example, bipolar plates of fuel cells or electrolysers or electrode plates of redox flow cells.
  • results of other test results that are relevant for the bipolar plate manufacture or process data for reinforcement learning are used as label or reward information.
  • the detailed tests can be integrated 100% inline or randomly if the process time is slower. In this case, only the amount of weld seams checked in detail in the training data set of the machine learning process is taken into account, which grows in the course of the bipolar plate production and gains in sharpness over time.
  • An essential manufacturing step in the manufacture of a fuel cell, in particular a polymer electrolyte fuel cell, is the manufacture of the bipolar plate, which takes place in particular by welding two or more metallic plates or sheets.
  • the weld seam produced during welding can be easily checked by means of a visual inspection.
  • This visual inspection is carried out automatically, however, in that the surface of the bipolar plate, in particular both flat sides of the bipolar plate, is recorded in the image with the aid of a camera and is fed to an image processing system of the automated evaluation system.
  • the evaluation system can process the image and, for example, use optical parameters to identify certain shapes and / or colors and / or brightnesses and / or contrasts and check whether these are within certain target values.
  • Such checking activities can be easily automated and implemented with the aid of evaluation software in the evaluation system.
  • it can also be checked whether a weld seam runs in a correct position relative to, for example, media-carrying channels, forms or structures, for example in the area of a gas distribution field.
  • the characteristics and structures of the welded bipolar plates can be identified in the image of the surface of the bipolar plate and used, for example, as a reference for the correct positioning of the weld seam.
  • dimensional deviations of the plate and its structures in particular the three-dimensional structures in the area of a gas distribution field, the formation of the openings in the area of the plate ends and the gas distribution field and the like, can also be checked in order to identify and sort out irregularities and to take corrective action in the area of the respective method step that potentially generates the deviation.
  • the image processing evaluation of the image does not make it possible, for example, to identify the correct execution of the characteristics or structures in the area of a gas distribution field with the required reliability, this can be due to a manufacturing error, so that such defects can also be detected with the help of the evaluation system the bipolar plate can be checked.
  • the bipolar plates found to be potentially “not in order” by the evaluation system are subjected to a detailed inspection as suspect plates, in which an inspection is carried out with significantly more effort, if necessary.
  • the examination effort can be reduced. For example, this means that an eddy current test is not carried out on every bipolar plate, but only on those suspect plates that were found to be suspicious during the automated visual inspection by the evaluation system.
  • the bipolar plates assessed as harmless during the automated image processing can thus be processed further without a further detailed examination, while after the detailed examination only those suspect plates are sorted out as rejects that actually do not meet the intended criteria and those suspect plates that still meet the criteria applied and can be fed back for further processing. Separation of a suspect plate as reject, which is unnecessary for actual reasons, is avoided, so that unnecessary costs are avoided.
  • the evaluation method used in the image processing of the image in the automated evaluation system allows the inspection quality to be increased when inspecting the fuel cell without significant additional effort, so that a cost-effective and reliable inspection of a fuel cell is made possible.
  • At least one non-destructive test in particular penetrant test, magnetic particle test, ultrasonic test, radiographic test and / or eddy current test, is carried out in the detailed test on the suspect plate in the area identified as potentially defective. A destruction of the suspicion plate this can be avoided during the detailed inspection. In the event that the detailed check should show that the suspicion plate does meet the requirement profile, the suspicion plate can be returned to the subsequent manufacturing steps as a regular bipolar plate. Unnecessary costs in the production of the fuel cell can thereby be avoided. At the same time, in comparison to a purely visual inspection, a measurement of the conditions present inside the suspect plate can take place, whereby a particularly reliable inspection is achieved. The detailed test can be carried out automatically and / or by a test person trained for this purpose.
  • the requirement profile can in particular relate to the tightness and position of a weld seam, the permeability and position of a structure in the gas distribution field, the arrangement of openings for the fuel gas supply, the design and position of seals and the like.
  • the evaluation system has a neural network trained on the basis of a large number of non-defective and defective bipolar plates depicting images of bipolar plates for identifying a suspect plate.
  • the neural network can be trained by machine learning so that the image evaluation of the images in the evaluation system can be improved. This makes it possible, in particular, to take into account a large number of different, possibly differently weighted parameters when evaluating the images.
  • the weighting of the parameters is automatically adjusted during the training of the neural network in such a way that, when evaluating the image used for the training, it is already known in advance whether a bipolar plate with this type of appearance is OK or not not, the correct images will be rated as “okay” or “not okay”. This makes it possible to automatically recognize complex damage patterns even with complex geometries. This will improve the quality of the exam.
  • the neural network is continuously trained with the results of the detailed check.
  • the detailed examination of the suspect plates leads to an additional gain in knowledge when evaluating the bipolar plates, with additional training data being permanently fed back to the evaluation system.
  • "Deep Learning” German: multilayered learning, deep learning or deep learning
  • Machine learning is a self-adaptive algorithm. Deep learning is a subset of machine learning and uses a number of hierarchical layers or a hierarchy of concepts to carry out the machine learning process.
  • two or more metal sheets are welded to produce the bipolar plate, with several images of a weld seam occurring during welding being created and evaluated and / or an image of the entire weld seam being created and evaluated after welding.
  • the image evaluation in the evaluation system can already take place during the welding process, so that melt zones that arise when the areas to be welded are melted can be taken into account in the evaluation.
  • the evaluation system preferably uses the results of the evaluation to adapt a process control of the welding.
  • the evaluation system can classify the possible defect on the basis of the criterion used for this purpose. This makes it possible to assign certain identified defects to a certain classification and, if necessary, to certain causes. If the defect of the suspect plate assigned to a certain class of defects can be assigned to a faulty process control during welding, it is possible, by feedback of the results of the evaluation by the evaluation system with the process control of the welding, to adapt the process control so that such defects can be avoided in the future.
  • the evaluation system can be part of a control loop for the process control of the welding, for example in order to carry out wear-related adaptations of control variables. As a result, manufacturing errors during welding can be recognized very early on and corrected automatically, so that unnecessary rejects and unnecessary costs are avoided.
  • Another embodiment relates to a test system for testing a bipolar plate of a fuel cell by means of the method according to the invention, with a camera for creating at least one image of a surface of the bipolar plate, an automated image processing-supported evaluation system for examining the image for possible defects and a switch that can be operated by the evaluation system for derivation a bipolar plate identified by the evaluation system as a potentially defective suspect plate to a detailed inspection station.
  • the test system is set up to carry out the method described above. As explained above with reference to the method, the test system is trained and further developed.
  • the bipolar plates found to be potentially “not in order” by the evaluation system are subjected to a detailed inspection as suspect plates, in which an inspection is carried out with significantly more effort, if necessary. However, since only the suspect plates are subjected to the detailed inspection in the detailed inspection station, the inspection effort can be reduced.
  • the bipolar plates assessed as harmless in the automated image processing in the evaluation system can thus be processed further without a further detailed check, while of the suspect plates after the detailed check only those are rejected as rejects that actually do not meet the intended criteria. Those suspect plates that still meet the criteria are sent back for further processing. An unnecessary separation of a suspect plate as scrap is avoided, so that unnecessary costs are avoided. Due to the evaluation method used in the image processing of the image in the automated evaluation system, the inspection can be The quality of the testing of the bipolar plate can be increased, so that a cost-effective and reliable test is made possible.
  • the evaluation system of the test system has a neural network trained on the basis of a large number of non-defective and defective bipolar plates depicting images of bipolar plates for identifying a suspect plate, the evaluation system having an interface that is coupled to the neural network and communicates with the detailed test station for feeding in the results of the detailed test for the purpose of continuous training of the neural network with the results of the detailed test.
  • the detailed inspection of the suspect plates in the detailed inspection station leads to an additional gain in knowledge when evaluating the bipolar plates, which is fed back to the evaluation system as training data via the interface of the evaluation system. This leads to reinforcement learning of the neural network, in particular as part of a deep learning structure with several intermediate layers of the neural network. As a result, the quality of the test will continue to improve over time.
  • the evaluation unit has an output port that can be coupled to a production unit for producing the bipolar plate for adapting a process control of the production unit as a function of the results of the evaluation by the evaluation system.
  • a production unit is, for example, a laser welding system, a punching system, a forming system and the like.
  • the evaluation system can classify the possible defect on the basis of the criterion used for this purpose. This makes it possible to assign certain identified defects to a certain classification and, if necessary, to certain causes.
  • the defect of the suspect plate assigned to a certain class of defects can be assigned to a faulty process control in the production unit, it is possible, by feeding back the results of the evaluation via the output port of the evaluation system to the production unit, to adapt the process control so that such defects can be avoided in the future.
  • the evaluation system can be part of a control loop for the process control of the production unit, for example in order to make adjustments to control variables due to wear. Manufacturing defects can thus be identified very early on and corrected automatically so that unnecessary rejects and unnecessary costs are avoided.
  • the detailed testing station preferably has at least one testing device for non-destructive testing, in particular for penetrant testing and / or magnetic particle testing and / or ultrasonic testing and / or radiographic testing and / or eddy current testing, of the suspect plate. Destruction of the suspicion plate during the detailed inspection can thus be avoided. In the event that the detailed inspection should show that the suspicion plate does meet the requirement profile, the suspicion plate can be returned to the subsequent positioning steps as a regular bipolar plate. This avoids unnecessary costs when setting up the components of a fuel cell. At the same time, in comparison to a purely visual inspection, a measurement of the conditions present inside the suspect plate can take place, whereby a particularly reliable inspection is achieved. The detailed inspection is carried out automatically, whereby an inspector can be dispensed with due to the constant improvement of the inspection quality.
  • the method according to the invention and the test system according to the invention are particularly suitable for testing bipolar plates for use in a fuel cell, in particular a polymer electrolyte fuel cell, the bipolar plate being formed from two or more metal sheets.
  • FIG. 2 a schematic perspective view of part of a test system for carrying out a test on a bipolar plate.
  • images 12 are provided, which in particular show the surface of a bipolar plate 14 after a welding process.
  • the images 12 are divided into two categories, for example, in which one category 16 is provided for images 12, which represent a bipolar plate 14 that is “OK”, and the other category 18 for Figures 12 are provided, which show a bipolar plate 14 that is “not in order” (NOK).
  • the category 18 provided for defective bipolar plates 14 can have several sub-categories, to which different causes for different defect patterns are assigned.
  • a neural network 20 is trained with the aid of the images 12 divided into categories 16, 18.
  • the neural network 20 is then automatically assigned the rating “OK” or “NOK” in a fourth step during a manufacturing process for the fuel cell 14.
  • the bipolar plates 14 found to be “in order” are fed to a further production step.
  • the suspicion plates assessed as potentially “not in order” are subjected in a fifth step to a detailed check 22 in order to confirm or correct the assessment given by the neural network 20 as “NOK”.
  • the detailed check 22 automatically produces a result in which, with a very high level of reliability, the neural network 20, at least as a case of doubt, still in the images 12 sorted under the “NOK” category as actually “NOK” or as “OK”. be classified. This corresponds to the state after the second step, in which training data were generated for the neural network 20.
  • These data automatically generated as a result of the detailed check 22 are fed back to the neural network 20 as additional training data in order to achieve reinforcement learning of the neural network 20.
  • the test system 24 can have a camera 26 which generates an image 12 of a surface of a, in particular welded, bipolar plate 14.
  • the camera 26 can also generate a plurality of images 12 or even a film sequence.
  • the bipolar plate 14 can be suitably illuminated for this purpose so that all relevant features can be seen in the illustration 12 and a possible defect 28 can be recognized by an evaluation system having the neural network 20 by means of an optical evaluation.
  • dimensional deviations of the plate and its structures in particular the three-dimensional structures in the area of a gas distribution field, the formation of the openings in the area of the plate ends and the gas distribution field and the like, can also be checked in order to identify and sort out irregularities and to take corrective action in the area of the respective method step that potentially generates the deviation.

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • General Physics & Mathematics (AREA)
  • Chemical & Material Sciences (AREA)
  • General Health & Medical Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Biochemistry (AREA)
  • Analytical Chemistry (AREA)
  • Immunology (AREA)
  • Pathology (AREA)
  • Theoretical Computer Science (AREA)
  • Evolutionary Computation (AREA)
  • Artificial Intelligence (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Signal Processing (AREA)
  • Electrochemistry (AREA)
  • Chemical Kinetics & Catalysis (AREA)
  • Quality & Reliability (AREA)
  • Data Mining & Analysis (AREA)
  • Computational Linguistics (AREA)
  • Biophysics (AREA)
  • Molecular Biology (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Biomedical Technology (AREA)
  • Sustainable Energy (AREA)
  • General Chemical & Material Sciences (AREA)
  • Sustainable Development (AREA)
  • Manufacturing & Machinery (AREA)
  • Acoustics & Sound (AREA)
  • Fuel Cell (AREA)
  • Mechanical Engineering (AREA)
  • Investigating Materials By The Use Of Optical Means Adapted For Particular Applications (AREA)
  • Battery Electrode And Active Subsutance (AREA)
  • Image Analysis (AREA)
  • Secondary Cells (AREA)

Abstract

Provision is made for a method for testing a bipolar plate of an electrochemical cell, in particular of a fuel cell, in which an image (12) of a surface of a bipolar plate (14) is created, the image (12) is investigated for possible defects (28) by an automated image processing-assisted evaluation system, and in the event of the evaluation system identifying an investigated bipolar plate (14) as a potentially defective suspect plate, a detailed test (22) is performed on a region of the suspect plate identified as potentially defective. By virtue of the evaluation methods applied in the automated evaluation system during the image processing of the image (12), it is possible to increase the test quality when testing a bipolar plate of an electrochemical cell, such as a fuel cell, without significant additional outlay, thereby allowing an inexpensive and reliable test.

Description

x Verfahren und Prüfanlage zum Prüfen einer Bipolarplatte einer elektrochemischen Zelle, insbesondere einer Brennstoffzelle x Method and test system for testing a bipolar plate of an electrochemical cell, in particular a fuel cell
Die Erfindung betrifft ein Verfahren zum Prüfen einer Bipolarplatte, insbesondere einer elektrochemischen Zelle wie einer Brennstoffzelle, sowie eine hierzu vorgesehene Prüfanlage, mit deren Hilfe eine Bipolarplatte einer elektrochemischen Zelle wie einer Brennstoffzelle während der Herstellung auf Defekte untersucht werden kann. The invention relates to a method for testing a bipolar plate, in particular an electrochemical cell such as a fuel cell, and a test system provided for this purpose, with the aid of which a bipolar plate of an electrochemical cell such as a fuel cell can be examined for defects during manufacture.
Ein Brennstoffzellenstack oder -Stapel beinhaltet in der Regel eine Vielzahl von Brennstoffzellen in einer Stapel-Konfiguration. Die jeweilige Brennstoffzelle umfasst einen Elektrolyten und Elektroden, die über leitfähige Platten kontaktiert werden. Im Falle einer Polymerelektrolytbrennstoffzelle, die im Niedertemperaturbereich betreiben wird, ist eine Polymermembran-Elektroden-Einheit vorhanden. Zur Trennung der ein- zelnen Polymerelektrolytbrennstoffzellen in einem Stack werden elektrisch leitfähige Bipolarplatten, meist aus Metall, verwendet. Sie dienen nicht nur zur elektrischen Kon- taktierung der Elektroden und leiten den elektrischen Strom zur benachbarten Zeile, sondern helfen auch, Brennstoffe und Kühlmedien zuzuführen und zu verteilen sowie Wärme und Reaktionsprodukte abzuführen. Dazu weisen Bipolarplatten üblicherweise Gasverieiierfeider auf, die aufgrund ihrer Struktur zu einer optimalen Verteilung der gasförmigen Brennstoffe, meist Wasserstoff und Sauerstoff, im Hinblick auf die Memb- ranoberflächen führen. Die Bipolarplatte kann dabei durch zwei oder mehrere dünn- wandige, miteinander verschweißte metallische Bleche hergestellt sein, welche die gewünschten Strömungspfade, Kühlkanäle und/oder Öffnungen ausbilden. A fuel cell stack or stack typically includes a plurality of fuel cells in a stack configuration. The respective fuel cell comprises an electrolyte and electrodes, which are contacted via conductive plates. In the case of a polymer electrolyte fuel cell that is operated in the low temperature range, a polymer membrane-electrode unit is present. Electrically conductive bipolar plates, mostly made of metal, are used to separate the individual polymer electrolyte fuel cells in a stack. They not only serve to make electrical contact with the electrodes and conduct the electrical current to the neighboring row, but also help to supply and distribute fuels and cooling media as well as to dissipate heat and reaction products. For this purpose, bipolar plates usually have gas distribution fields which, due to their structure, lead to an optimal distribution of the gaseous fuels, mostly hydrogen and oxygen, with regard to the membrane surfaces. The bipolar plate can be produced by two or more thin-walled metal sheets welded to one another, which form the desired flow paths, cooling channels and / or openings.
Aus US 2019/0296379 A1 ist es bekannt, eine Brennstoffzelle zu prüfen, indem ver- schiedene Spannungen angelegt und ausgewertet werden. From US 2019/0296379 A1 it is known to test a fuel cell by applying and evaluating different voltages.
Die US 2019/0340747 A1 offenbart ein Güteüberwachungssystem und -verfahren im Bereich einer Brennstoffzellenfertigungslinie. Die US 2013/0230072 A1 beschreibt ein Verfahren und eine Vorrichtung zur Fehlerer- kennung an Brennstoffzellenkomponenten, insbesondere Bipolarplatten. US 2019/0340747 A1 discloses a quality monitoring system and method in the area of a fuel cell production line. US 2013/0230072 A1 describes a method and a device for error detection on fuel cell components, in particular bipolar plates.
Die DE 10393237 B4 offenbart ein Verfahren zum Detektieren elektrischer Defekte in Membranelektrodenanordnungen. DE 10393237 B4 discloses a method for detecting electrical defects in membrane electrode arrangements.
Die DE 3809221 A1 beschreibt ein Verfahren und eine Vorrichtung zum Detektieren von Fehlstellen, insbesondere Rissen und/oder Einschnürungen, an Preßteilen oder anderen Werkstücken. DE 3809221 A1 describes a method and a device for detecting defects, in particular cracks and / or constrictions, on pressed parts or other workpieces.
Die DE102015221 697 B3 offenbart eine Anordnung zur Bestimmung der Oberflä- chenbeschaffenheit von Bauteiloberflächen, insbesondere von Defekten. DE102015221 697 B3 discloses an arrangement for determining the surface quality of component surfaces, in particular of defects.
Die DE 102016211 449 A1 beschreibt ein Prüfsystem mit einer tragbaren Prüfeinheit und ein Verfahren zum Prüfen von Bauteilen. DE 102016211 449 A1 describes a test system with a portable test unit and a method for testing components.
Die DE 102009059765 A1 offenbart ein Verfahren zur Fierstellung einer Bipolarplat- te. Die Bipolarplatte wird schrittweise aus einer ersten und einer zweiten Platte gebil- det, wobei nach einem jeden Schritt ein Istmuster aus einer Vielzahl von Linien durch Projektion auf einem den jeweiligen Schritt repräsentierenden Zwischenprodukt er- zeugt und mittels einer Kamera aufgenommen wird. Anschließend wird mit einem Re- ferenzmuster verglichen. DE 102009059765 A1 discloses a method for positioning a bipolar plate. The bipolar plate is formed step by step from a first and a second plate, after each step an actual pattern of a large number of lines is generated by projection onto an intermediate product representing the respective step and recorded by means of a camera. It is then compared with a reference pattern.
Es besteht ein ständiges Bedürfnis, die Prüfung von Komponenten einer elektroche- mischen Zelle wie einer Brennstoffzelle kostengünstig und verlässlich durchzuführen, bevor diese in einer elektrochemischen Zelle, wie einer Brennstoffzelle oder einem Brennstoffzellenstapel, mit weiteren Bauteilen verbaut werden. There is a constant need to test components of an electrochemical cell such as a fuel cell in a cost-effective and reliable manner before they are installed with further components in an electrochemical cell such as a fuel cell or a fuel cell stack.
Es ist die Aufgabe der Erfindung Maßnahmen aufzuzeigen, die eine kostengünstige und verlässliche Prüfung der Komponenten einer elektrochemischen Zelle, insbeson- dere Brennstoffzelle, ermöglichen. It is the object of the invention to identify measures which enable the components of an electrochemical cell, in particular a fuel cell, to be tested inexpensively and reliably.
Die Lösung der Aufgabe erfolgt durch ein Verfahren mit den Merkmalen des An- spruchs 1 sowie eine Prüfanlage mit den Merkmalen des Anspruchs 7. Bevorzugte Ausgestaltungen der Erfindung sind in den Unteransprüchen und der nachfolgenden Beschreibung angegeben, die jeweils einzeln oder in Kombination einen Aspekt der Erfindung darstellen können. The object is achieved by a method with the features of claim 1 and a test system with the features of claim 7. Preferred Refinements of the invention are specified in the subclaims and the following description, each of which can represent an aspect of the invention individually or in combination.
Eine Ausführungsform betrifft ein Verfahren zum Prüfen einer Bipolarplatte einer elekt- rochemischen Zell, insbesondere einer Brennstoffzelle, bei dem eine Abbildung einer Oberfläche einer Bipolarplatte erstellt wird, die Abbildung von einem automatisierten bildverarbeitungsunterstützten Auswertesystem auf mögliche Defekte untersucht wird und in dem Fall, dass das Auswertesystem eine untersuchte Bipolarplatte als potenti- ell defekte Verdachts-Platte identifiziert, eine Detailprüfung eines als potentiell defekt identifizierten Bereichs der Verdachts-Platte durchgeführt wird, wobei das Auswerte- system ein, anhand einer Vielzahl von nicht defekte und defekte Bipolarplatten dar- stellenden Abbildungen von Bipolarplatten trainiertes neuronales Netzwerk zur Identi- fizierung einer Verdachts-Platte aufweist und das neuronale Netzwerk mit den Ergeb- nissen der Detailprüfung fortlaufend trainiert wird, wobei das neuronale Netzwerk ein Verstärkungslernen durchführt. One embodiment relates to a method for testing a bipolar plate of an electrochemical cell, in particular a fuel cell, in which an image of a surface of a bipolar plate is created, the image is examined for possible defects by an automated image processing-assisted evaluation system and in the event that the evaluation system an examined bipolar plate is identified as a potentially defective suspect plate, a detailed examination of an area of the suspect plate identified as potentially defective is carried out, the evaluation system displaying images of bipolar plates using a large number of non-defective and defective bipolar plates has a trained neural network for identifying a suspect plate and the neural network is continuously trained with the results of the detailed check, the neural network performing reinforcement learning.
Siehe hierzu auch den Artikel mit dem Titel „So funktioniert Reinforcement Learning“, vom 11.04.2019, unter: https://www.alexanderthamm.com/de/blog/einfach-erklaert-so-funktioniert- reinforcement-learning/ See also the article entitled "How Reinforcement Learning Works", dated April 11, 2019, at: https://www.alexanderthamm.com/de/blog/einfach-erklaert-so-funktioniert- reinforcement-learning /
Unter einer Bipolarplatte wird im Sinne der vorliegenden Erfindung ein Bauteil einer elektrochemischen Zelle verstanden, welches die Funktionen einer Elektrode und ei- nes Strömungsleitelements bereitstellt. Dabei kann es sich beispielsweise um Bipolar- platten von Brennstoffzellen oder Elektrolyseuren oder auch Elektrodenplatten von Redox-Flow-Zellen handeln. In the context of the present invention, a bipolar plate is understood to be a component of an electrochemical cell which provides the functions of an electrode and a flow guide element. This can be, for example, bipolar plates of fuel cells or electrolysers or electrode plates of redox flow cells.
Im Kontext einer Bipolarplattenprüfung werden für das Verstärkungslernen insbeson- dere Ergebnisse anderer, bei der Bipolarplattenherstellung relevanter Prüfergebnisse oder Prozessdaten für das Verstärkungslernen als Label- oder Belohnungsinformation hinzugezogen. So werden bevorzugt Ergebnisse einer Dichtheitsprüfung, Inline- Prozessdaten des Schweißprozesses, eine thermografische Detailprüfung, eine topo- graphische Prüfung unter Bestimmung der Güte einer Schweißnaht oder Röntgen/CT- Informationen zu entsprechenden Schweißnähten einbezogen. Die Detailprüfungen können inline zu 100% oder auch stichprobenartig bei langsamerer Prozesszeit ein- gebunden werden. In diesem Fall wird nur die im Detail geprüfte Menge an Schweiß- nähten im Trainingsdatensatz des Machine-Learning-Verfahrens berücksichtigt, die im Lauf der Bipolarplattenproduktion wächst und an Schärfe mit der Zeit hinzugewinnt. In the context of a bipolar plate test, for reinforcement learning, in particular the results of other test results that are relevant for the bipolar plate manufacture or process data for reinforcement learning are used as label or reward information. The results of a leak test, inline process data of the welding process, a thermographic detailed test, a topographical test determining the quality of a weld seam or X-ray / CT Information on corresponding welds included. The detailed tests can be integrated 100% inline or randomly if the process time is slower. In this case, only the amount of weld seams checked in detail in the training data set of the machine learning process is taken into account, which grows in the course of the bipolar plate production and gains in sharpness over time.
Ein wesentlicher Herstellungsschritt bei der Herstellung einer Brennstoffzelle, insbe- sondere einer Polymerelektrolytbrennstoffzelle, ist die Fertigung der Bipolarplatte, die insbesondere durch ein Verschweißen von zwei oder mehreren metallischen Platten oder Blechen erfolgt. Die beim Verschweißen entstehende Schweißnaht kann gut durch eine Sichtprüfung überprüft werden. Diese Sichtprüfung wird jedoch automati- siert durchgeführt, indem die Oberfläche der Bipolarplatte, insbesondere beide Flach- seiten der Bipolarplatte, mit Hilfe einer Kamera in der Abbildung festgehalten wird und einer Bildverarbeitung des automatisierten Auswertesystems zugeführt wird. Das Auswertesystem kann die Abbildung verarbeiten und beispielswiese anhand optischer Parameter bestimmte Formen und/oder Farben und/oder Helligkeiten und/oder Kon- traste identifizieren und überprüfen, ob diese innerhalb bestimmter Sollwerte liegen. Derartige Überprüfungstätigkeiten sind gut automatisierbar und mit Hilfe einer Auswer- tesoftware des Auswertesystems umsetzbar. Hierbei kann zudem überprüft werden, ob eine Schweißnaht in einer korrekten Relativlage zu, beispielsweise medienführen- de Kanäle ausbildenden, Ausprägungen oder Strukturen, beispielsweise im Bereich eines Gasverteilerfeldes, verläuft. Die Ausprägungen und Strukturen der geschweiß- ten Bipolarplatten können hierbei in der Abbildung der Oberfläche der Bipolarplatte identifiziert und beispielsweise als Referenz für die korrekte Positionierung der Schweißnaht verwendet werden. An essential manufacturing step in the manufacture of a fuel cell, in particular a polymer electrolyte fuel cell, is the manufacture of the bipolar plate, which takes place in particular by welding two or more metallic plates or sheets. The weld seam produced during welding can be easily checked by means of a visual inspection. This visual inspection is carried out automatically, however, in that the surface of the bipolar plate, in particular both flat sides of the bipolar plate, is recorded in the image with the aid of a camera and is fed to an image processing system of the automated evaluation system. The evaluation system can process the image and, for example, use optical parameters to identify certain shapes and / or colors and / or brightnesses and / or contrasts and check whether these are within certain target values. Such checking activities can be easily automated and implemented with the aid of evaluation software in the evaluation system. Here it can also be checked whether a weld seam runs in a correct position relative to, for example, media-carrying channels, forms or structures, for example in the area of a gas distribution field. The characteristics and structures of the welded bipolar plates can be identified in the image of the surface of the bipolar plate and used, for example, as a reference for the correct positioning of the weld seam.
Neben Defekten betreffend die Schweißverbindung an einer Bipolarplatte können wei- terhin Dimensionsabweichungen der Platte und deren Strukturen, insbesondere der dreidimensionalen Strukturen im Bereich eines Gasverteilerfeldes, der Ausbildung der Öffnungen im Bereich der Plattenenden und des Gasverteilerfeldes und dergleichen überprüft werden, um Unregelmäßigkeiten zu identifizieren, auszusortieren und im Be- reich des jeweiligen, die Abweichung potentiell generierenden Verfahrensschritts kor- rigierend einzugreifen. Wenn es bei der bildverarbeitenden Auswertung der Abbildung jedoch nicht möglich sein sollte, beispielsweise die korrekte Ausführung der Ausprägungen oder Strukturen im Bereich eines Gasverteilerfeldes mit der erforderlichen Sicherheit zu identifizieren, kann dies auf einen Herstellungsfehler zurückzuführen sein, so dass mit Hilfe des Auswertesystems auch derartige Defekte der Bipolarplatte überprüft werden können. In addition to defects relating to the welded connection on a bipolar plate, dimensional deviations of the plate and its structures, in particular the three-dimensional structures in the area of a gas distribution field, the formation of the openings in the area of the plate ends and the gas distribution field and the like, can also be checked in order to identify and sort out irregularities and to take corrective action in the area of the respective method step that potentially generates the deviation. However, if the image processing evaluation of the image does not make it possible, for example, to identify the correct execution of the characteristics or structures in the area of a gas distribution field with the required reliability, this can be due to a manufacturing error, so that such defects can also be detected with the help of the evaluation system the bipolar plate can be checked.
Die von dem Auswertesystem als potentiell „nicht in Ordnung“ befundenen Bipolarplat- ten werden als Verdachts-Platten der Detailprüfung unterzogen, in der mit gegebenen- falls deutlich höherem Aufwand eine Prüfung durchgeführt wird. Da jedoch nur die Verdachts-Platten der Detailprüfung unterzogen werden, kann der Prüfungsaufwand reduziert werden. Beispielsweise wird dadurch nicht bei jeder Bipolarplatte eine Wir- belstromprüfung durchgeführt, sondern nur bei denjenigen Verdachts-Platten, die bei der automatisierten Sichtprüfung durch das Auswertesystem als verdächtig aufgefal- len sind. Hierbei wird die Erkenntnis ausgenutzt, dass eine optisch nahezu perfekt aussehende Schweißnaht erfahrungsgemäß unproblematisch ist, während bei einer nicht perfekt aussehenden Schweißnaht zwar eine noch ausreichende, insbesondere ausreichend dichte, Verbindung vorliegen kann, aber dies nicht in allen Fällen durch eine reine Sichtprüfung garantiert werden kann. Die bei der automatisierten Bildverar- beitung als unbedenklich bewerteten Bipolarplatten können dadurch ohne eine weitere Detailprüfung weiterverarbeitet werden, während von den Verdachts-Platten nach der Detailprüfung nur diejenigen als Ausschuss aussortiert werden, die den vorgesehenen Kriterien tatsächlich nicht entsprechen und diejenigen Verdachts-Platten, die den an- gelegten Kriterien doch noch entsprechen, wieder der Weiterverarbeitung zugeführt werden können. Eine aus tatsächlichen Gründen unnötige Aussonderung einer Ver- dachts-Platte als Ausschuss wird vermieden, so dass unnötige Kosten vermieden sind. Durch die bei der Bildverarbeitung der Abbildung in dem automatisierten Aus- wertesystem angewendeten Auswerteverfahren kann ohne signifikanten Zusatzauf- wand die Prüfungsqualität beim Prüfen der Brennstoffzelle erhöht werden, so dass ei- ne kostengünstige und verlässliche Prüfung einer Brennstoffzelle ermöglicht ist. The bipolar plates found to be potentially “not in order” by the evaluation system are subjected to a detailed inspection as suspect plates, in which an inspection is carried out with significantly more effort, if necessary. However, since only the suspect plates are subjected to the detailed examination, the examination effort can be reduced. For example, this means that an eddy current test is not carried out on every bipolar plate, but only on those suspect plates that were found to be suspicious during the automated visual inspection by the evaluation system. This makes use of the knowledge that experience shows that a weld seam that looks almost perfect is not a problem, while a weld seam that does not look perfect may still have a sufficient, in particular sufficiently tight, connection, but this cannot be guaranteed in all cases by a pure visual inspection . The bipolar plates assessed as harmless during the automated image processing can thus be processed further without a further detailed examination, while after the detailed examination only those suspect plates are sorted out as rejects that actually do not meet the intended criteria and those suspect plates that still meet the criteria applied and can be fed back for further processing. Separation of a suspect plate as reject, which is unnecessary for actual reasons, is avoided, so that unnecessary costs are avoided. The evaluation method used in the image processing of the image in the automated evaluation system allows the inspection quality to be increased when inspecting the fuel cell without significant additional effort, so that a cost-effective and reliable inspection of a fuel cell is made possible.
Insbesondere wird in der Detailprüfung mindestens eine zerstörungsfreie Prüfung, insbesondere Eindringprüfung, Magnetpulverprüfung, Ultraschallprüfung, Durchstrah- lungsprüfung und/oder Wirbelstromprüfung, an der Verdachts-Platte in dem als poten- tiell defekt identifizierten Bereich durchgeführt. Eine Zerstörung der Verdachts-Platte bei der Detailprüfung kann dadurch vermieden werden. In dem Fall, dass die De- tailprüfung ergeben sollte, dass die Verdachts-Platte doch dem Anforderungsprofil ge- nügt, kann die Verdachts-Platte als reguläre Bipolarplatte wieder den nachfolgenden Herstellungsschritten zugeführt werden. Unnötige Kosten bei der Herstellung der Brennstoffzelle können dadurch vermieden werden. Gleichzeitig kann im Vergleich zu einer reinen Sichtprüfung eine Messung von im Inneren der Verdachts-Platte vorlie- genden Verhältnissen erfolgen, wodurch eine besonders zuverlässige Prüfung erreicht wird. Die Detailprüfung kann automatisiert und/oder durch eine hierzu ausgebildete Prüfperson durchgeführt werden. In particular, at least one non-destructive test, in particular penetrant test, magnetic particle test, ultrasonic test, radiographic test and / or eddy current test, is carried out in the detailed test on the suspect plate in the area identified as potentially defective. A destruction of the suspicion plate this can be avoided during the detailed inspection. In the event that the detailed check should show that the suspicion plate does meet the requirement profile, the suspicion plate can be returned to the subsequent manufacturing steps as a regular bipolar plate. Unnecessary costs in the production of the fuel cell can thereby be avoided. At the same time, in comparison to a purely visual inspection, a measurement of the conditions present inside the suspect plate can take place, whereby a particularly reliable inspection is achieved. The detailed test can be carried out automatically and / or by a test person trained for this purpose.
Das Anforderungsprofil kann dabei insbesondere die Dichtigkeit und Position einer Schweißnaht, die Durchlässigkeit und Position einer Struktur im Gasverteilerfeld, die Anordnung von Öffnungen zur Brenngaszufuhr, die Ausbildung und Position von Dich- tungen und dergleichen betreffen. The requirement profile can in particular relate to the tightness and position of a weld seam, the permeability and position of a structure in the gas distribution field, the arrangement of openings for the fuel gas supply, the design and position of seals and the like.
Erfindungsgemäß weist das Auswertesystem ein anhand einer Vielzahl von nicht de- fekte und defekte Bipolarplatten darstellenden Abbildungen von Bipolarplatten trainier- tes neuronales Netzwerk zur Identifizierung einer Verdachts-Platte auf. Das neuronale Netzwerk kann durch Maschinenlernen trainiert werden, so dass die Bildauswertung der Abbildungen in dem Auswertesystem verbessert werden kann. Dadurch ist es ins- besondere möglich, eine Vielzahl verschiedener, gegebenenfalls unterschiedlich stark gewichteter Parameter bei der Auswertung der Abbildungen zu berücksichtigen. Dabei wird die Gewichtung der Parameter während des Trainings des neuronalen Netzwer- kes automatisiert derart angepasst, dass bei der Bildauswertung der für das Training verwendeten Abbildung, bei denen im Vorhinein bereits bekannt ist, ob eine Bipolar- platte mit einem derartigen Aussehen in Ordnung ist oder nicht, die korrekten Abbil- dungen als „in Ordnung“ beziehungsweise „nicht in Ordnung“ bewerten werden. Dies ermöglicht es, auch bei komplexen Geometrien komplexe Schadmuster automatisiert zu erkennen. Die Prüfungsqualität wird dadurch verbessert werden. According to the invention, the evaluation system has a neural network trained on the basis of a large number of non-defective and defective bipolar plates depicting images of bipolar plates for identifying a suspect plate. The neural network can be trained by machine learning so that the image evaluation of the images in the evaluation system can be improved. This makes it possible, in particular, to take into account a large number of different, possibly differently weighted parameters when evaluating the images. The weighting of the parameters is automatically adjusted during the training of the neural network in such a way that, when evaluating the image used for the training, it is already known in advance whether a bipolar plate with this type of appearance is OK or not not, the correct images will be rated as “okay” or “not okay”. This makes it possible to automatically recognize complex damage patterns even with complex geometries. This will improve the quality of the exam.
Das neuronale Netzwerk wird erfindungsgemäß mit den Ergebnissen der Detailprü- fung fortlaufend trainiert. Die Detailprüfung der Verdachts-Platten führt zu einem zu- sätzlichen Erkenntnisgewinn bei der Auswertung der Bipolarplatten, wobei zusätzliche Trainingsdaten permanent an das Auswertesystem rückgeführt werden. Dies führt zu einem Verstärkungslernen des neuronalen Netzwerks, insbesondere als Teil einer Deeplearning-Struktur mit mehreren Zwischenschichten des neuronalen Netzwerks. „Deep Learning“ (deutsch: mehrschichtiges Lernen, tiefes Lernen oder tiefgehendes Lernen) bezeichnet eine Methode des maschinellen Lernens, die künstliche neuronale Netze mit zahlreichen Zwischenschichten, sogenannten „hidden layers“, zwischen Eingabeschicht und Ausgabeschicht einsetzt. Dadurch bildet sich eine umfangreiche innere Struktur heraus. Es ist eine Methode der Informationsverarbeitung. Maschinel- les Lernen ist ein selbstadaptiver Algorithmus. Deep Learning ist eine Teilmenge des maschinellen Lernens und nutzt eine Reihe hierarchischer Schichten bzw. eine Hie- rarchie von Konzepten, um den Prozess des maschinellen Lernens durchzuführen.According to the invention, the neural network is continuously trained with the results of the detailed check. The detailed examination of the suspect plates leads to an additional gain in knowledge when evaluating the bipolar plates, with additional training data being permanently fed back to the evaluation system. this leads to reinforcement learning of the neural network, in particular as part of a deep learning structure with several intermediate layers of the neural network. "Deep Learning" (German: multilayered learning, deep learning or deep learning) describes a method of machine learning that uses artificial neural networks with numerous intermediate layers, so-called "hidden layers", between the input layer and the output layer. This creates an extensive internal structure. It is a method of information processing. Machine learning is a self-adaptive algorithm. Deep learning is a subset of machine learning and uses a number of hierarchical layers or a hierarchy of concepts to carry out the machine learning process.
Die Prüfungsgüte wird dadurch im Laufe der Zeit immer weiter verbessert. Das Ver- stärkungslernen kann kontinuierlich oder inkrementeil erfolgen. As a result, the quality of the test will continue to improve over time. Reinforcement learning can take place continuously or incrementally.
Insbesondere erfolgt zur Herstellung der Bipolarplatte ein Verschweißen von zwei oder mehr Blechen, wobei während des Schweißens mehrere Abbildungen einer beim Schweißen entstehenden Schweißnaht erstellt und ausgewertet werden und/oder nach dem Schweißen eine Abbildung der gesamten Schweißnaht erstellt und ausge- wertet wird. Die Bildauswertung in dem Auswertesystem kann bereits während des Schweißvorgangs erfolgen, so dass auch beim Aufschmelzen der zu verschweißen- den Bereiche entstehende Schmelzzonen bei der Auswertung berücksichtigt werden können. Es ist aber auch möglich, die Auswertung erst nach Abschluss des Schweiß- vorgangs vorzunehmen, wodurch die bei der Auswertung zu berücksichtigenden Da- tenmengen, insbesondere die Anzahl an Abbildungen, reduziert werden kann. Zudem ist es möglich eine Beleuchtung der Bipolarplatte und eine Erstellung der Abbildung mit Hilfe einer Kamera unabhängig vom Schweißprozess und/oder unabhängig von beim Schweißen entstehenden Lichteffekten und/oder reversibler Wärmedehnungsef- fekte durchzuführen. In particular, two or more metal sheets are welded to produce the bipolar plate, with several images of a weld seam occurring during welding being created and evaluated and / or an image of the entire weld seam being created and evaluated after welding. The image evaluation in the evaluation system can already take place during the welding process, so that melt zones that arise when the areas to be welded are melted can be taken into account in the evaluation. However, it is also possible to carry out the evaluation only after the welding process has been completed, as a result of which the data quantities to be taken into account in the evaluation, in particular the number of images, can be reduced. In addition, it is possible to illuminate the bipolar plate and create the image with the aid of a camera independently of the welding process and / or independently of light effects and / or reversible thermal expansion effects occurring during welding.
Vorzugsweise erfolgt mit den Ergebnissen der Auswertung durch das Auswertesystem eine Anpassung einer Prozesssteuerung des Schweißens. Das Auswertesystem kann insbesondere in dem Fall, dass eine Bipolarplatte als „nicht in Ordnung“ bewertet wur- de, anhand des hierzu herangezogenen Kriteriums eine Klassifizierung des möglichen Defekts vornehmen. Dies ermöglicht es, bestimmte identifizierte Defekte einer be- stimmten Klassifizierung und gegebenenfalls bestimmten Ursachen zuzuordnen. Wenn der, einer bestimmten Klasse von Defekten zugeordnete Defekt der Verdachts- Platte einer fehlerhaften Prozessführung beim Schweißen zugeordnet werden kann, ist es möglich, durch die Rückkoppelung der Ergebnisse der Auswertung durch das Auswertesystem mit der Prozessführung des Schweißens die Prozessführung so an- zupassen, dass derartige Defekte zukünftig vermieden werden. Das Auswertesystem kann Teil eines Regelkreises für die Prozessführung des Schweißens sein, beispiels- weise um verschleißbedingte Anpassungen von Regelgrößen vorzunehmen. Herstel- lungsfehler beim Schweißen können dadurch sehr frühzeitig erkannt und automatisiert korrigiert werden, so dass unnötiger Ausschuss und unnötige Kosten vermieden sind. The evaluation system preferably uses the results of the evaluation to adapt a process control of the welding. In particular in the event that a bipolar plate was assessed as “not in order”, the evaluation system can classify the possible defect on the basis of the criterion used for this purpose. This makes it possible to assign certain identified defects to a certain classification and, if necessary, to certain causes. If the defect of the suspect plate assigned to a certain class of defects can be assigned to a faulty process control during welding, it is possible, by feedback of the results of the evaluation by the evaluation system with the process control of the welding, to adapt the process control so that such defects can be avoided in the future. The evaluation system can be part of a control loop for the process control of the welding, for example in order to carry out wear-related adaptations of control variables. As a result, manufacturing errors during welding can be recognized very early on and corrected automatically, so that unnecessary rejects and unnecessary costs are avoided.
Eine weitere Ausführungsform betrifft eine Prüfanlage zum Prüfen einer Bipolarplatte einer Brennstoffzelle mittels des erfindungsgemäßen Verfahrens, mit einer Kamera zur Erstellung mindestens einer Abbildung einer Oberfläche der Bipolarplatte, einem automatisierten bildverarbeitungsunterstützten Auswertesystem zur Untersuchung der Abbildung auf mögliche Defekte und einer von dem Auswertesystem betätigbaren Weiche zur Ableitung einer von dem Auswertesystem als potentiell defekte Verdachts- Platte identifizierte Bipolarplatte an eine Detailprüfungsstation. Die Prüfanlage ist zur Durchführung des vorstehend beschriebenen Verfahrens hergerichtet. Die Prüfanlage ist dabei, wie vorstehend anhand des Verfahrens erläutert, aus- und weitergebildet. Another embodiment relates to a test system for testing a bipolar plate of a fuel cell by means of the method according to the invention, with a camera for creating at least one image of a surface of the bipolar plate, an automated image processing-supported evaluation system for examining the image for possible defects and a switch that can be operated by the evaluation system for derivation a bipolar plate identified by the evaluation system as a potentially defective suspect plate to a detailed inspection station. The test system is set up to carry out the method described above. As explained above with reference to the method, the test system is trained and further developed.
Die von dem Auswertesystem als potentiell „nicht in Ordnung“ befundenen Bipolarplat- ten werden als Verdachts-Platten der Detailprüfung unterzogen, in der mit gegebenen- falls deutlich höherem Aufwand eine Prüfung durchgeführt wird. Da jedoch nur die Verdachts-Platten der Detailprüfung in der Detailprüfungsstation unterzogen werden, kann der Prüfungsaufwand reduziert werden. Die bei der automatisierten Bildverarbei- tung in dem Auswertesystem als unbedenklich bewerteten Bipolarplatten können dadurch ohne eine weitere Detailprüfung weiterverarbeitet werden, während von den Verdachts-Platten nach der Detailprüfung nur diejenigen als Ausschuss aussortiert werden, die den vorgesehenen Kriterien tatsächlich nicht entsprechen. Diejenigen Verdachts-Platten, die den angelegten Kriterien doch noch entsprechen, werden wie- der der Weiterverarbeitung zugeführt. Eine unnötige Aussonderung einer Verdachts- Platte als Ausschuss wird vermieden, so dass unnötige Kosten vermieden sind. Durch die bei der Bildverarbeitung der Abbildung in dem automatisierten Auswertesystem angewendeten Auswerteverfahren kann ohne signifikanten Zusatzaufwand die Prü- fungsqualität beim Prüfen der Bipolarplatte erhöht werden, so dass eine kostengünsti- ge und verlässliche Prüfung ermöglicht ist. The bipolar plates found to be potentially “not in order” by the evaluation system are subjected to a detailed inspection as suspect plates, in which an inspection is carried out with significantly more effort, if necessary. However, since only the suspect plates are subjected to the detailed inspection in the detailed inspection station, the inspection effort can be reduced. The bipolar plates assessed as harmless in the automated image processing in the evaluation system can thus be processed further without a further detailed check, while of the suspect plates after the detailed check only those are rejected as rejects that actually do not meet the intended criteria. Those suspect plates that still meet the criteria are sent back for further processing. An unnecessary separation of a suspect plate as scrap is avoided, so that unnecessary costs are avoided. Due to the evaluation method used in the image processing of the image in the automated evaluation system, the inspection can be The quality of the testing of the bipolar plate can be increased, so that a cost-effective and reliable test is made possible.
Das Auswertesystem der erfindungsgemäßen Prüfanlage weist ein anhand einer Viel- zahl von nicht defekte und defekte Bipolarplatten darstellenden Abbildungen von Bipo- larplatten trainiertes neuronales Netzwerk zur Identifizierung einer Verdachts-Platte auf, wobei das Auswertesystem eine mit dem neuronalen Netzwerk gekoppelte, mit der Detailprüfungsstation kommunizierende Schnittstelle zur Einspeisung der Ergeb- nisse der Detailprüfung zum Zwecke des fortlaufendenTrainings des neuronalen Netzwerks mit den Ergebnissen der Detailprüfung aufweist. Die Detailprüfung der Verdachts-Platten in der Detailprüfungsstation führt zu einem zusätzlichen Erkennt- nisgewinn bei der Auswertung der Bipolarplatten, der über die Schnittstelle des Aus- wertesystems als Trainingsdaten an das Auswertesystem rückgeführt wird. Dies führt zu einem Verstärkungslernen des neuronalen Netzwerks, insbesondere als Teil einer Deeplearning-Struktur mit mehreren Zwischenschichten des neuronalen Netzwerks. Die Prüfungsgüte wird dadurch im Laufe der zeit immer weiter verbessert. The evaluation system of the test system according to the invention has a neural network trained on the basis of a large number of non-defective and defective bipolar plates depicting images of bipolar plates for identifying a suspect plate, the evaluation system having an interface that is coupled to the neural network and communicates with the detailed test station for feeding in the results of the detailed test for the purpose of continuous training of the neural network with the results of the detailed test. The detailed inspection of the suspect plates in the detailed inspection station leads to an additional gain in knowledge when evaluating the bipolar plates, which is fed back to the evaluation system as training data via the interface of the evaluation system. This leads to reinforcement learning of the neural network, in particular as part of a deep learning structure with several intermediate layers of the neural network. As a result, the quality of the test will continue to improve over time.
Insbesondere weist die Auswerteeinheit einen mit einer Produktionseinheit zur Her- stellung der Bipolarplatte koppelbaren Ausgangsport zur Anpassung einer Prozess- steuerung der Produktionseinheit in Abhängigkeit der Ergebnisse der Auswertung durch das Auswertesystem auf. Eine solche Produktionseinheit ist beispielsweise eine Laserschweißanlage, eine Stanzanlage, eine Umformanlage und dergleichen. Das Auswertesystem kann insbesondere in dem Fall, dass eine Bipolarplatte als „nicht in Ordnung“ bewertet wurde, anhand des hierzu herangezogenen Kriteriums eine Klassi- fizierung des möglichen Defekts vornehmen. Dies ermöglicht es, bestimmte identifi- zierte Defekte einer bestimmten Klassifizierung und gegebenenfalls bestimmten Ursa- chen zuzuordnen. Wenn der einer bestimmten Klasse von Defekten zugeordnete De- fekt der Verdachts-Platte einer fehlerhaften Prozessführung in der Produktionseinheit zugeordnet werden kann, ist es möglich, durch die Rückkoppelung der Ergebnisse der Auswertung über den Ausgangsport des Auswertesystems mit der Produktionseinheit die Prozessführung so anzupassen, dass derartige Defekte zukünftig vermieden wer- den. Das Auswertesystem kann Teil eines Regelkreises für die Prozessführung der Produktionseinheit sein, beispielsweise um verschleißbedingte Anpassungen von Re- gelgrößen vorzunehmen. Herstellungsfehler können dadurch sehr frühzeitig erkannt und automatisiert korrigiert werden, so dass unnötiger Ausschuss und unnötige Kos- ten vermieden werden. In particular, the evaluation unit has an output port that can be coupled to a production unit for producing the bipolar plate for adapting a process control of the production unit as a function of the results of the evaluation by the evaluation system. Such a production unit is, for example, a laser welding system, a punching system, a forming system and the like. In particular in the event that a bipolar plate was assessed as “not in order”, the evaluation system can classify the possible defect on the basis of the criterion used for this purpose. This makes it possible to assign certain identified defects to a certain classification and, if necessary, to certain causes. If the defect of the suspect plate assigned to a certain class of defects can be assigned to a faulty process control in the production unit, it is possible, by feeding back the results of the evaluation via the output port of the evaluation system to the production unit, to adapt the process control so that such defects can be avoided in the future. The evaluation system can be part of a control loop for the process control of the production unit, for example in order to make adjustments to control variables due to wear. Manufacturing defects can thus be identified very early on and corrected automatically so that unnecessary rejects and unnecessary costs are avoided.
Vorzugsweise weist die Detailprüfungsstation mindestens ein Prüfgerät zur zerstö- rungsfreien Prüfung, insbesondere zur Eindringprüfung und/oder Magnetpulverprü- fung und/oder Ultraschallprüfung und/oder Durchstrahlungsprüfung und/oder Wirbels- tromprüfung, der Verdachts-Platte auf. Eine Zerstörung der Verdachts-Platte bei der Detailprüfung kann dadurch vermieden werden. In dem Fall, dass die Detailprüfung ergeben sollte, dass die Verdachts-Platte doch dem Anforderungsprofil genügt, kann die Verdachts-Platte als reguläre Bipolarplatte wieder den nachfolgenden Fierstel- lungsschritten zugeführt werden. Unnötige Kosten bei der Fierstellung der Komponen- ten einer Brennstoffzelle können dadurch vermieden werden. Gleichzeitig kann im Vergleich zu einer reinen Sichtprüfung eine Messung von im Inneren der Verdachts- Platte vorliegenden Verhältnissen erfolgen, wodurch eine besonders zuverlässige Prü- fung erreicht wird. Die Detailprüfung wird automatisiert durchgeführt, wobei aufgrund der stetigen Verbesserung der Prüfgüte auf eine Prüfperson verzichtet werden kann. The detailed testing station preferably has at least one testing device for non-destructive testing, in particular for penetrant testing and / or magnetic particle testing and / or ultrasonic testing and / or radiographic testing and / or eddy current testing, of the suspect plate. Destruction of the suspicion plate during the detailed inspection can thus be avoided. In the event that the detailed inspection should show that the suspicion plate does meet the requirement profile, the suspicion plate can be returned to the subsequent positioning steps as a regular bipolar plate. This avoids unnecessary costs when setting up the components of a fuel cell. At the same time, in comparison to a purely visual inspection, a measurement of the conditions present inside the suspect plate can take place, whereby a particularly reliable inspection is achieved. The detailed inspection is carried out automatically, whereby an inspector can be dispensed with due to the constant improvement of the inspection quality.
Das erfindungsgemäße Verfahren und die erfindungsgemäße Prüfanlage sind insbe- sondere zum Prüfen von Bipolarplatten zur Verwendung in einer Brennstoffzelle, ins- besondere einer Polymerelektrolytbrennstoffzelle, geeignet, wobei die Bipolarplatte aus zwei oder mehreren metallischen Blechen gebildet ist. The method according to the invention and the test system according to the invention are particularly suitable for testing bipolar plates for use in a fuel cell, in particular a polymer electrolyte fuel cell, the bipolar plate being formed from two or more metal sheets.
Nachfolgend wird die Erfindung unter Bezugnahme auf die anliegenden Zeichnungen anhand bevorzugter Ausführungsbeispiele exemplarisch erläutert, wobei die nachfol- gend dargestellten Merkmale sowohl jeweils einzeln als auch in Kombination einen Aspekt der Erfindung darstellen können. Es zeigen: In the following, the invention is explained by way of example with reference to the attached drawings using preferred exemplary embodiments, the features shown below being able to represent an aspect of the invention both individually and in combination. Show it:
Fig. 1 : einen schematischen Ablauf für die Durchführung einer Prüfung einer Bipolar- platte, und 1: a schematic sequence for carrying out a test of a bipolar plate, and
Fig. 2: eine schematische perspektivische Ansicht eines Teils einer Prüfanlage für die die Durchführung einer Prüfung bei einer Bipolarplatte. Bei dem in Fig. 1 dargestellten Ablauf 10 werden in einem ersten Schritt mehrere Ab- bildungen 12 bereitgestellt, die insbesondere die Oberfläche einer Bipolarplatte 14 nach einem Schweißvorgang darstellen. In einem zweiten Schritt werden die Abbil- dungen 12 beispielsweise in zwei Kategorien eingeteilt, bei der die eine Kategorie 16 für Abbildungen 12 vorgesehen ist, welche eine Bipolarplatte 14 darstellen, die „in Ordnung“ (OK) ist, und die andere Kategorie 18 für Abbildungen 12 vorgesehen ist, welche eine Bipolarplatte 14 darstellen, die „nicht in Ordnung“ (NOK) ist. Insbesonde- re kann die für defekte Bipolarplatten 14 vorgesehene Kategorie 18 mehrere Unterka- tegorien aufweisen, welchen verschiedenen Ursachen für verschiedene Defektmuster zugeordnet sind. In einem dritten Schritt wird ein neuronales Netzwerk 20 mit Hilfe der in die Kategorien 16, 18 eingeteilten Abbildungen 12 trainiert. Das neuronale Netz- werk 20 wird dann in einem vierten Schritt während eines Herstellungsprozesses der Brennstoffzelle 14 selbständig die Bewertung „OK“ oder „NOK“ vergeben. Die als „in Ordnung“ befundenen Bipolarplatten 14 werden einem weiteren Fertigungsschritt zu- geführt. Die als potentiell „nicht in Ordnung“ bewerteten Verdachts-Platten werden in einem fünften Schritt einer Detailprüfung 22 unterzogen, um die vom neuronalen Netzwerk 20 vergebene Bewertung als „NOK“ zu bestätigen oder zu korrigieren. Bei der Detailprüfung 22 entsteht automatisch ein Ergebnis, bei dem mit einer sehr hohen Verlässlichkeit von dem neuronalen Netzwerk 20 zumindest als Zweifelsfall noch in die unter der „NOK“-Kategorie einsortierten Abbildungen 12 als tatsächlich „NOK“ o- der doch als „OK“ klassifiziert werden. Dies entspricht dem Zustand nach dem zweiten Schritt, in dem Trainingsdaten für das neuronale Netzwerk 20 generiert wurden. Diese als Ergebnis der Detailprüfung 22 automatisch erzeugten Daten werden als zusätzli- che Trainingsdaten an das neuronale Netzwerk 20 rückgekoppelt, um ein Verstär- kungslernen des neuronalen Netzwerks 20 zu erreichen. 2: a schematic perspective view of part of a test system for carrying out a test on a bipolar plate. In the sequence 10 shown in FIG. 1, in a first step several images 12 are provided, which in particular show the surface of a bipolar plate 14 after a welding process. In a second step, the images 12 are divided into two categories, for example, in which one category 16 is provided for images 12, which represent a bipolar plate 14 that is “OK”, and the other category 18 for Figures 12 are provided, which show a bipolar plate 14 that is “not in order” (NOK). In particular, the category 18 provided for defective bipolar plates 14 can have several sub-categories, to which different causes for different defect patterns are assigned. In a third step, a neural network 20 is trained with the aid of the images 12 divided into categories 16, 18. The neural network 20 is then automatically assigned the rating “OK” or “NOK” in a fourth step during a manufacturing process for the fuel cell 14. The bipolar plates 14 found to be “in order” are fed to a further production step. The suspicion plates assessed as potentially “not in order” are subjected in a fifth step to a detailed check 22 in order to confirm or correct the assessment given by the neural network 20 as “NOK”. The detailed check 22 automatically produces a result in which, with a very high level of reliability, the neural network 20, at least as a case of doubt, still in the images 12 sorted under the “NOK” category as actually “NOK” or as “OK”. be classified. This corresponds to the state after the second step, in which training data were generated for the neural network 20. These data automatically generated as a result of the detailed check 22 are fed back to the neural network 20 as additional training data in order to achieve reinforcement learning of the neural network 20.
Wie in Fig. 2 dargestellt ist, kann die Prüfanlage 24 eine Kamera 26 aufweisen, die von einer Oberfläche einer, insbesondere geschweißten, Bipolarplatte 14 eine Abbil- dung 12 erzeugt. Die Kamera 26 kann auch eine Mehrzahl von Abbildungen 12 oder sogar eine Filmsequenz erzeugen. Die Bipolarplatte 14 kann hierzu geeignet ausge- leuchtet sein, damit in der Abbildung 12 alle relevanten Merkmale zu sehen sind und ein möglicher Defekt 28 von einem das neuronale Netzwerk 20 aufweisenden Auswer- tesystem durch eine optische Auswertung erkannt werden kann. Neben Defekten betreffend die Schweißverbindung an einer Bipolarplatte können wei- terhin Dimensionsabweichungen der Platte und deren Strukturen, insbesondere der dreidimensionalen Strukturen im Bereich eines Gasverteilerfeldes, der Ausbildung der Öffnungen im Bereich der Plattenenden und des Gasverteilerfeldes und dergleichen überprüft werden, um Unregelmäßigkeiten zu identifizieren, auszusortieren und im Be- reich des jeweiligen, die Abweichung potentiell generierenden Verfahrensschritts kor- rigierend einzugreifen. As shown in FIG. 2, the test system 24 can have a camera 26 which generates an image 12 of a surface of a, in particular welded, bipolar plate 14. The camera 26 can also generate a plurality of images 12 or even a film sequence. The bipolar plate 14 can be suitably illuminated for this purpose so that all relevant features can be seen in the illustration 12 and a possible defect 28 can be recognized by an evaluation system having the neural network 20 by means of an optical evaluation. In addition to defects relating to the welded connection on a bipolar plate, dimensional deviations of the plate and its structures, in particular the three-dimensional structures in the area of a gas distribution field, the formation of the openings in the area of the plate ends and the gas distribution field and the like, can also be checked in order to identify and sort out irregularities and to take corrective action in the area of the respective method step that potentially generates the deviation.
Bezugszeichenliste List of reference symbols
10 Ablauf 10 process
12 Abbildung 14 Bipolarplatte 16 eine Kategorie 18 andere Kategorie 20 neuronales Netzwerk 22 Detailprüfung 24 Prüfanlage 26 Kamera 28 möglicher Defekt 12 Figure 14 Bipolar plate 16 one category 18 another category 20 neural network 22 detailed inspection 24 test system 26 camera 28 possible defect

Claims

Patentansprüche Claims
1 . Verfahren zum Prüfen einer Bipolarplatte einer elektrochemischen Zelle, insbe- sondere einer Brennstoffzelle, bei dem eine Abbildung (12) einer Oberfläche einer Bi- polarplatte (14) erstellt wird, die Abbildung (12) von einem automatisierten bildverarbeitungsunterstützten Auswer- tesystem auf mögliche Defekte (28) untersucht wird und in dem Fall, dass das Auswertesystem eine untersuchte Bipolarplatte (14) als potenti- ell defekte Verdachts-Platte identifiziert, eine Detailprüfung (22) eines als potentiell de- fekt identifizierten Bereichs der Verdachts-Platte durchgeführt wird, wobei das Auswertesystem ein, anhand einer Vielzahl von nicht defekte und defekte Bipolarplatten (14) darstellenden Abbildungen (12) von Bipolarplatten (14) trainiertes neuronales Netzwerk (20) zur Identifizierung einer Verdachts-Platte aufweist und das neuronale Netzwerk (20) mit den Ergebnissen der Detailprüfung (22) fortlaufend trainiert wird, wobei das neuronale Netzwerk (20) ein Verstärkungslernen durchführt. 1 . Method for testing a bipolar plate of an electrochemical cell, in particular a fuel cell, in which an image (12) of a surface of a bipolar plate (14) is created, the image (12) from an automated image processing-assisted evaluation system for possible defects ( 28) is examined and, in the event that the evaluation system identifies an examined bipolar plate (14) as a potentially defective suspect plate, a detailed check (22) of an area of the suspect plate identified as potentially defective is carried out Evaluation system has a neural network (20) trained on the basis of a large number of non-defective and defective bipolar plates (14) depicting images (12) of bipolar plates (14) to identify a suspect plate and the neural network (20) with the results of the detailed test (22) is continuously trained, the neural network (20) performing reinforcement learning t.
2. Verfahren nach Anspruch 1 , bei dem in der Detailprüfung (22) mindestens eine zerstörungsfreie Prüfung an der Verdachts-Platte in dem als potentiell defekt identifi- zierten Bereich durchgeführt wird. 2. The method according to claim 1, in which in the detailed test (22) at least one non-destructive test is carried out on the suspect plate in the area identified as potentially defective.
3. Verfahren nach Anspruch 2, bei dem als zerstörungsfreie Prüfung eine Ein- dringprüfung und/oder eine Magnetpulverprüfung und/oder eine Ultraschallprüfung und/oder eine Durchstrahlungsprüfung und/oder eine Wirbelstromprüfung an der Ver- dachts-Platte in dem als potentiell defekt identifizierten Bereich durchgeführt wird. 3. The method according to claim 2, in which the non-destructive test is a penetration test and / or a magnetic particle test and / or an ultrasonic test and / or a radiographic test and / or an eddy current test on the suspect plate in the area identified as potentially defective is carried out.
4. Verfahren nach einem der Ansprüche 1 bis 3, bei dem zur Herstellung der Bipo- larplatte (14) ein Verschweißen von zwei oder mehreren Blechen erfolgt, wobei wäh- rend des Schweißens mehrere Abbildungen (12) einer beim Schweißen entstehenden Schweißnaht erstellt und ausgewertet werden. 4. The method as claimed in one of claims 1 to 3, in which two or more metal sheets are welded to produce the bipolar plate (14), with several images (12) of a weld seam occurring during welding being created and evaluated during welding will.
5. Verfahren nach einem der Ansprüche 1 bis 4, bei dem zur Herstellung der Bipo- larplatte (14) nach dem Schweißen eine Abbildung (12) der gesamten Schweißnaht erstellt und ausgewertet wird. 5. The method according to any one of claims 1 to 4, in which, to produce the bipolar plate (14) after welding, an image (12) of the entire weld seam is created and evaluated.
6. Verfahren nach Anspruch 4 oder 5, bei dem mit den Ergebnissen der Auswer- tung durch das Auswertesystem eine Anpassung einer Prozesssteuerung des Schweißens erfolgt. 6. The method according to claim 4 or 5, in which the results of the evaluation by the evaluation system are used to adapt a process control of the welding.
7. Prüfanlage zur Durchführung eines Verfahrens nach einem der Ansprüche 1 bis 6, mit einer Kamera (26) zur Erstellung mindestens einer Abbildung (12) einer Oberfläche der Bipolarplatte (14), einem automatisierten bildverarbeitungsunterstützten Auswertesystem zur Untersu- chung der Abbildung (12) auf mögliche Defekte (28) und einer von dem Auswertesystem betätigbaren Weiche zur Ableitung einer von dem Auswertesystem als potentiell defekte Verdachts-Platte identifizierte Bipolarplatte an eine Detailprüfungsstation, wobei das Auswertesystem ein, anhand einer Vielzahl von nicht defekte und defekte Bipolarplatten (14) darstellenden Abbildungen (12) von Bipolarplatten (14) trainiertes neuronales Netzwerk (20) zur Identifizierung einer Verdachts-Platte aufweist, wobei das Auswertesystem eine mit dem neuronalen Netzwerk (20) gekoppelte, mit der De- tailprüfungsstation kommunizierende Schnittstelle zur Einspeisung der Ergebnisse der Detailprüfung (22) zum Zwecke des fortlaufenden Trainings des neuronalen Netz- werks (20) mit den Ergebnissen der Detailprüfung (22) aufweist. 7. Test system for carrying out a method according to one of claims 1 to 6, with a camera (26) for creating at least one image (12) of a surface of the bipolar plate (14), an automated image processing-assisted evaluation system for examining the image (12) for possible defects (28) and a switch that can be operated by the evaluation system to derive a bipolar plate identified by the evaluation system as a potentially defective suspect plate to a detailed inspection station, the evaluation system showing an image based on a large number of non-defective and defective bipolar plates (14) (12) has a neural network (20) trained by bipolar plates (14) for identifying a suspect plate, the evaluation system having an interface coupled to the neural network (20) and communicating with the detail checking station for feeding in the results of the detailed checking (22 ) for the purpose of continuous training of the neural network (20) with the results of the detailed test (22).
8. Prüfanlage nach Anspruch 7, dadurch gekennzeichnet, dass die Auswerteein- heit mindestens einen mit mindestens einer Produktionseinheit, insbesondere einer Schweißanlage und/oder einer Stanzanlage und/oder einer Umformanlage, zur Her- stellung der Bipolarplatte (14) koppelbaren Ausgangsport zur Anpassung einer Pro- zesssteuerung der Produktionseinheit in Abhängigkeit der Ergebnisse der Auswertung durch das Auswertesystem aufweist. 8. Testing system according to claim 7, characterized in that the evaluation unit has at least one output port that can be coupled to at least one production unit, in particular a welding system and / or a punching system and / or a forming system, for producing the bipolar plate (14) for adapting a Has process control of the production unit as a function of the results of the evaluation by the evaluation system.
9. Prüfanlage nach Anspruch 8, dadurch gekennzeichnet, dass die mindestens eine Produktionseinheit eine Schweißanlage und/oder eine Stanzanlage und/oder ei- ne Umformanlage ist. 9. Testing system according to claim 8, characterized in that the at least one production unit is a welding system and / or a stamping system and / or a forming system.
10. Prüfanlage nach einem der Ansprüche 7 bis 9, dadurch gekennzeichnet, dass die Detailprüfungsstation mindestens ein Prüfgerät zur zerstörungsfreien Prüfung, ins- besondere zur Eindringprüfung und/oder Magnetpulverprüfung und/oder Ultraschall- prüfung und/oder Durchstrahlungsprüfung und/oder Wirbelstromprüfung, der Ver- dachts-Platte aufweist. 10. Testing system according to one of claims 7 to 9, characterized in that the detailed testing station has at least one testing device for non-destructive testing, in particular for penetrant testing and / or magnetic particle testing and / or ultrasonic testing and / or radiographic testing and / or eddy current testing, the Ver - has dachts plate.
EP21717314.5A 2020-03-20 2021-02-26 Method and test facility for testing a bipolar plate of an electrochemical cell, in particular of a fuel cell Pending EP4122042A1 (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
DE102020107779.3A DE102020107779A1 (en) 2020-03-20 2020-03-20 Method and test system for testing a bipolar plate of a fuel cell
PCT/DE2021/100195 WO2021185404A1 (en) 2020-03-20 2021-02-26 Method and test facility for testing a bipolar plate of an electrochemical cell, in particular of a fuel cell

Publications (1)

Publication Number Publication Date
EP4122042A1 true EP4122042A1 (en) 2023-01-25

Family

ID=75438539

Family Applications (1)

Application Number Title Priority Date Filing Date
EP21717314.5A Pending EP4122042A1 (en) 2020-03-20 2021-02-26 Method and test facility for testing a bipolar plate of an electrochemical cell, in particular of a fuel cell

Country Status (6)

Country Link
EP (1) EP4122042A1 (en)
JP (1) JP7438382B2 (en)
KR (1) KR20220130755A (en)
CN (1) CN115088125A (en)
DE (1) DE102020107779A1 (en)
WO (1) WO2021185404A1 (en)

Families Citing this family (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114267844B (en) * 2021-11-09 2024-02-27 深圳市氢瑞燃料电池科技有限公司 System and method for producing fuel cell polar plate
KR20230110189A (en) * 2022-01-14 2023-07-21 주식회사 엘지에너지솔루션 Apparatus for monitoring and operating method of the same
CN114577816A (en) * 2022-01-18 2022-06-03 广州超音速自动化科技股份有限公司 Hydrogen fuel bipolar plate detection method
DE102022109188B3 (en) 2022-04-14 2023-08-03 Schaeffler Technologies AG & Co. KG Fuel cell stack and method of assembling a fuel cell stack
WO2024049194A1 (en) * 2022-08-31 2024-03-07 주식회사 엘지에너지솔루션 Artificial intelligence model-based abnormality diagnosis method, and abnormality diagnosis device and factory monitoring system using same

Family Cites Families (25)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DE3809221A1 (en) 1988-03-18 1989-09-28 Roth Electric Gmbh METHOD FOR DETECTING DEFECTS IN PRESSING PARTS OR OTHER WORKPIECES, AND DEVICE FOR IMPLEMENTING THE METHOD
JP2968442B2 (en) * 1994-09-26 1999-10-25 川崎重工業株式会社 Evaluation system for welding defects
JP3936095B2 (en) 1999-03-31 2007-06-27 株式会社東芝 Fuel cell
US7179553B2 (en) 2002-09-06 2007-02-20 General Motors Corporation Method for detecting electrical defects in membrane electrode assemblies
JP2007018176A (en) 2005-07-06 2007-01-25 Sharp Corp Learning device, learning method, learning program, recording medium, and device and method for pattern recognition
JP5005218B2 (en) 2005-12-28 2012-08-22 愛知機械工業株式会社 Inspection apparatus and inspection method
DE102009059765A1 (en) 2009-12-21 2011-06-22 Daimler AG, 70327 Bipolar plate producing method, involves producing actual pattern from lines by projection on intermediate products, and recording actual pattern by camera, where actual pattern is compared with reference pattern
US9784625B2 (en) 2010-11-30 2017-10-10 Bloom Energy Corporation Flaw detection method and apparatus for fuel cell components
JP2013167596A (en) 2012-02-17 2013-08-29 Honda Motor Co Ltd Defect inspection device, defect inspection method, and program
JP6237263B2 (en) 2014-01-24 2017-11-29 日産自動車株式会社 Manufacturing method of fuel cell
CN104833925A (en) * 2015-05-07 2015-08-12 昆山弗尔赛能源有限公司 Fuel cell bipolar plate detection method and system based on machine vision
DE102015221697B3 (en) 2015-11-05 2017-02-23 Fraunhofer-Gesellschaft zur Förderung der angewandten Forschung e.V. Arrangement for determining the surface quality of component surfaces
DE102016211449A1 (en) 2016-06-27 2017-12-28 Bayerische Motoren Werke Aktiengesellschaft Test system and method for testing components
KR102557181B1 (en) * 2017-01-10 2023-07-18 케이엘에이 코포레이션 Diagnostic methods for faults captured by optical tools and classifiers
JP7021728B2 (en) 2017-05-12 2022-02-17 株式会社神戸製鋼所 Automatic welding system, welding control method, and machine learning model
US10809635B2 (en) * 2017-11-20 2020-10-20 Taiwan Semiconductor Manufacturing Company, Ltd. Defect inspection method and defect inspection system
CN108320278A (en) * 2018-01-09 2018-07-24 北京百度网讯科技有限公司 Product defects detect localization method, device, equipment and computer-readable medium
US11069030B2 (en) * 2018-03-22 2021-07-20 Adobe, Inc. Aesthetics-guided image enhancement
JP6990610B2 (en) 2018-03-23 2022-02-03 本田技研工業株式会社 Fuel cell current leak inspection method
CN108631727B (en) * 2018-03-26 2019-08-09 河北工业大学 A kind of solar panel defect identification method based on convolutional neural networks
US10769770B2 (en) 2018-05-07 2020-09-08 Cummins Enterprise Llc Quality monitoring system and quality monitoring method for fuel cell manufacturing line and quality monitoring system for manufacturing line
DE102018214307A1 (en) * 2018-08-23 2020-02-27 Friedrich-Alexander-Universität Erlangen-Nürnberg System and method for quality inspection in the manufacture of individual parts
CN109598721B (en) * 2018-12-10 2021-08-31 广州市易鸿智能装备有限公司 Defect detection method and device for battery pole piece, detection equipment and storage medium
CN110132237A (en) * 2019-05-05 2019-08-16 四川省地质工程勘察院 A kind of method of urban ground deformation disaster EARLY RECOGNITION
CN110473806A (en) * 2019-07-13 2019-11-19 河北工业大学 The deep learning identification of photovoltaic cell sorting and control method and device

Also Published As

Publication number Publication date
JP7438382B2 (en) 2024-02-26
KR20220130755A (en) 2022-09-27
WO2021185404A1 (en) 2021-09-23
JP2023514753A (en) 2023-04-07
DE102020107779A1 (en) 2021-09-23
CN115088125A (en) 2022-09-20

Similar Documents

Publication Publication Date Title
WO2021185404A1 (en) Method and test facility for testing a bipolar plate of an electrochemical cell, in particular of a fuel cell
DE102013109915B4 (en) Method and device for checking an inspection system for detecting surface defects
DE102021127389A1 (en) AUTOMATED TESTING AND VERIFICATION OF ELECTRIC MOTOR WELDING QUALITY
DE102015204800B3 (en) Method and device for quality evaluation of a component produced by means of an additive manufacturing method
DE102020002826A1 (en) Method for monitoring a laser welding process using optical coherence tomography
DE112012003688T5 (en) Fuel cell inspection method and inspection device
EP3839494B1 (en) Method for characterization of a weld seam
WO2021165380A1 (en) Method for analysing a weld during laser welding of workpieces
DE102019005331A1 (en) Method for testing a weld seam
DE102008046692A1 (en) Destruction-free testing of welding seam, which is produced by a friction stir welding process using a friction stir tool between two joining partners, comprises locally introducing heat in an area of the welding seam
EP4288932B1 (en) Method and device for evaluating a weld quality of a weld joint between end sections of two conductor elements
DE102011052562B4 (en) Method for producing a gas-tight fuel cell and testing device therefor
Tantrapiwat Spot welding defect detection using synthetic image dataset on convolutional neural networks
DE102014218269A1 (en) Method and device for producing a resin welding article
DE102008038332A1 (en) Robotic welding process and assembly for e.g. automobile assembly uses module-logged electrical and optical welding parameters
DE202013011975U1 (en) Apparatus for the production of steel barrels
DE102004028607B4 (en) Apparatus and method for damage-free testing of the connection quality of stud welding joints
DE102021127016A1 (en) Process signal reconstruction and anomaly detection in laser machining processes
DE102018109816B3 (en) A method for obtaining at least one significant feature in a series of components of the same type and method for classifying a component of such a series
DE102018203269A1 (en) Method for the qualitative analysis of a spot weld, welding device and use of acoustic signals
EP4187673A1 (en) Method for inspecting lithium secondary battery
KR102400012B1 (en) Method for diagnosing seam weld quality
DE102022205117A1 (en) Method for testing the suitability of a metal surface for a welding process and method for welding components
DE102005041465B3 (en) Detecting welding irregularities e.g., for industry, involves using measurement device to receive value pairs of weld current and voltage synchronously from welding point
EP4155852A1 (en) Processor circuit and method for monitoring a joining process of workpieces in a joining machine and joining machine

Legal Events

Date Code Title Description
STAA Information on the status of an ep patent application or granted ep patent

Free format text: STATUS: UNKNOWN

STAA Information on the status of an ep patent application or granted ep patent

Free format text: STATUS: THE INTERNATIONAL PUBLICATION HAS BEEN MADE

PUAI Public reference made under article 153(3) epc to a published international application that has entered the european phase

Free format text: ORIGINAL CODE: 0009012

STAA Information on the status of an ep patent application or granted ep patent

Free format text: STATUS: REQUEST FOR EXAMINATION WAS MADE

17P Request for examination filed

Effective date: 20221020

AK Designated contracting states

Kind code of ref document: A1

Designated state(s): AL AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HR HU IE IS IT LI LT LU LV MC MK MT NL NO PL PT RO RS SE SI SK SM TR

DAV Request for validation of the european patent (deleted)
DAX Request for extension of the european patent (deleted)
P01 Opt-out of the competence of the unified patent court (upc) registered

Effective date: 20230523