WO2019102892A1 - 検査支援システム、学習装置、及び判定装置 - Google Patents
検査支援システム、学習装置、及び判定装置 Download PDFInfo
- Publication number
- WO2019102892A1 WO2019102892A1 PCT/JP2018/041847 JP2018041847W WO2019102892A1 WO 2019102892 A1 WO2019102892 A1 WO 2019102892A1 JP 2018041847 W JP2018041847 W JP 2018041847W WO 2019102892 A1 WO2019102892 A1 WO 2019102892A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- determination
- result
- unit
- learning
- determination result
- Prior art date
Links
- 238000007689 inspection Methods 0.000 title claims abstract description 63
- 238000012360 testing method Methods 0.000 claims abstract description 95
- 230000005540 biological transmission Effects 0.000 claims description 16
- 230000001066 destructive effect Effects 0.000 claims description 2
- 238000009659 non-destructive testing Methods 0.000 abstract description 2
- 238000000034 method Methods 0.000 description 6
- 230000035515 penetration Effects 0.000 description 4
- 230000005855 radiation Effects 0.000 description 4
- 238000003466 welding Methods 0.000 description 4
- 238000013528 artificial neural network Methods 0.000 description 3
- 239000006185 dispersion Substances 0.000 description 3
- 239000000463 material Substances 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 238000011179 visual inspection Methods 0.000 description 3
- 229910000831 Steel Inorganic materials 0.000 description 2
- 238000010276 construction Methods 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 239000010959 steel Substances 0.000 description 2
- 239000000126 substance Substances 0.000 description 2
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000004590 computer program Methods 0.000 description 1
- 238000012790 confirmation Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 230000010365 information processing Effects 0.000 description 1
- 239000006249 magnetic particle Substances 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 239000002245 particle Substances 0.000 description 1
- 230000002787 reinforcement Effects 0.000 description 1
- 239000002893 slag Substances 0.000 description 1
- WFKWXMTUELFFGS-UHFFFAOYSA-N tungsten Chemical compound [W] WFKWXMTUELFFGS-UHFFFAOYSA-N 0.000 description 1
- 229910052721 tungsten Inorganic materials 0.000 description 1
- 239000010937 tungsten Substances 0.000 description 1
Images
Classifications
-
- 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
- G01N29/4445—Classification of defects
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M99/00—Subject matter not provided for in other groups of this subclass
-
- 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
- G01N23/04—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 forming images of the material
-
- 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
- G01N23/083—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 the radiation being X-rays
-
- 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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/60—Type of objects
- G06V20/64—Three-dimensional objects
Definitions
- the present invention relates to an inspection support system for supporting nondestructive inspection of an object, and a learning apparatus and determination apparatus that can be used for the inspection support system.
- RT Radiographic Testing
- Patent Document 1 As a technique for supporting such nondestructive inspection, a technique for detecting a welded portion of a steel pipe is known (see, for example, Patent Document 1).
- the welding portion detection method described in Patent Document 1 extracts a video signal of the inner surface of a pipe using a television camera while rotating the steel pipe in the circumferential direction, and the pipe surface having a value unique to the pipe type from the obtained video signal.
- the image feature quantity is extracted, and a weld part and a base material part are distinguished by a neural network which learns and stores the inner face image feature quantity of the pipe type to be detected in advance, and the weld part is detected.
- the operator selects the signal waveform feature to be learned, and further causes the neural network learning device to input and learn the marking signal of the welding portion or the base material portion.
- the work is easy to belong to the person.
- the learning efficiency of the neural network learning device depends on the ability and the amount of work of the operator to be learned, there is a limit to the burden reduction of the operator.
- the present invention has been made in view of these circumstances, and an object thereof is to provide a technique for improving the efficiency of nondestructive inspection of an object.
- an inspection support system is an inspection support system for supporting nondestructive inspection of an object, and the pass / fail is determined based on the result of the nondestructive test of the object.
- the determination apparatus determines a pass / fail by the determination algorithm based on the test result acquisition unit that acquires the result of the nondestructive test of the object and the nondestructive test result of the object acquired by the test result acquisition unit.
- the learning device is received by the determination result receiving unit that receives the final determination result by the inspector and the result of the nondestructive test of the object corresponding to the determination result from the plurality of determination devices, and the determination result receiving unit And a providing unit that provides the plurality of determination devices with the determination algorithm learned by the learning unit.
- the nondestructive inspection of the object it is possible to automatically perform the pass / fail determination and sort out those suspected of failing prior to the final determination by the visual inspection of the inspector. Therefore, the efficiency and accuracy of nondestructive inspection can be improved.
- the number of inspectors can be significantly reduced, the burden on individual inspectors can be significantly reduced, and labor costs can be reduced.
- variation in the determination result by each inspector's capability etc. can be suppressed, the precision of nondestructive inspection can be improved.
- the learning efficiency of the judgment algorithm can be obtained because a much larger amount and a wide variety of information can be collected and used for learning the judgment algorithm than when the judgment algorithm is learned independently at each plant construction site. Also, the learning speed can be greatly improved, and the accuracy of the determination algorithm can be dramatically improved.
- the judgment result transmission unit acquires the corrected judgment result, and transmits it to the learning device together with the result of the nondestructive test of the object corresponding to the judgment result.
- the determination algorithm can be learned so as to correct the erroneous determination, the accuracy of the determination can be efficiently improved.
- the determination result transmission unit acquires the determination result by the inspector on the result of the nondestructive test of the object for which the determination unit could not determine the pass / fail, and the learning device together with the result of the nondestructive test of the object corresponding to the determination result It may be sent to
- the number of cases that can be determined by the determination algorithm can be increased, so that the accuracy of the determination can be efficiently improved.
- the determination apparatus may further include a learning unit that learns a determination algorithm based on the final determination result by the inspector and the result of the nondestructive test of the object corresponding to the determination result.
- the learning of the determination algorithm can be independently advanced in each of the determination devices, so that the accuracy of the determination can be efficiently improved.
- This apparatus is a target corresponding to the final judgment result by the inspector who confirmed the judgment result by the judgment apparatus from a plurality of judgment apparatuses which judge pass / fail based on the result of the nondestructive test of the object
- a determination result receiving unit that receives the result of the nondestructive test of the object, and a learning unit that learns a determination algorithm used for the determination of pass / fail in the plurality of determination devices based on the information received by the determination result receiving unit;
- a providing unit that provides the plurality of determination devices with the determination algorithm learned by the learning unit.
- the determination result receiving unit may receive the corrected determination result and the nondestructive test result of the object corresponding to the determination result when the determination result by the determination apparatus is corrected by the inspector.
- the determination algorithm can be learned so as to correct the erroneous determination, the accuracy of the determination can be efficiently improved.
- the determination result receiving unit receives the determination result by the inspector on the result of the nondestructive test of the object for which the determination device can not determine the pass / fail, and the result of the nondestructive test of the object corresponding to the determination result. Good.
- the number of cases that can be determined by the determination algorithm can be increased, so that the accuracy of the determination can be efficiently improved.
- Yet another aspect of the present invention is a determination apparatus.
- This apparatus uses a test result acquisition unit for acquiring the nondestructive test results of the object, and a determination algorithm used to determine pass or fail based on the nondestructive test results of the object acquired by the test result acquisition unit.
- a determination unit that determines the pass / fail by the determination algorithm provided by the learning device for learning the learning result, a determination result presentation unit that presents the determination result by the determination unit to an inspector who performs nondestructive inspection of the object;
- a determination result transmission unit that acquires a final determination result by the inspector who has confirmed the determination result by the unit and transmits the result to the learning device together with the result of the nondestructive test of the object corresponding to the determination result.
- the nondestructive inspection of the object it is possible to automatically perform the pass / fail determination and sort out those suspected of failing prior to the final determination by the visual inspection of the inspector. Therefore, the efficiency and accuracy of nondestructive inspection can be improved.
- the number of inspectors can be significantly reduced, the burden on individual inspectors can be significantly reduced, and labor costs can be reduced.
- variation in the determination result by each inspector's capability etc. can be suppressed, the precision of nondestructive inspection can be improved.
- the information processing apparatus may further include a learning unit that learns a determination algorithm based on the final determination result by the inspector and the result of the nondestructive test of the object corresponding to the determination result.
- the learning of the determination algorithm can be independently advanced in each of the determination devices, so that the accuracy of the determination can be efficiently improved.
- FIG. 1 shows the entire configuration of a test support system according to the embodiment.
- An inspection support system 1 for supporting nondestructive inspection of an object includes a plurality of plants 3 based on a plant 3 for producing chemical products and industrial products and information collected from a plurality of plants 3.
- a learning device 4 is provided to learn a determination algorithm 8 used to determine the success or failure of the nondestructive inspection.
- Each of the plants 3 includes an inspection object 10 such as a welded portion of piping installed in the plant 3, a test apparatus 20 for performing a nondestructive test of the inspection object 10, and a non-inspection object 10 by the test apparatus 20.
- the determination algorithm 8 determines the pass / fail, and a determination device 30 is provided which presents the determination result to the inspector who carries out the nondestructive inspection of the object.
- a determination device 30 is provided which presents the determination result to the inspector who carries out the nondestructive inspection of the object.
- Each plant 3 and the learning device 4 are connected by the Internet 2.
- the determination device 30 transmits the final determination result by the inspector who has confirmed the determination result by the determination device 30 to the learning device 4 together with the result of the nondestructive test of the object corresponding to the determination result.
- the learning device 4 includes a determination result receiving unit 5, a learning unit 6, a providing unit 7, and a determination algorithm 8.
- These hardware components are realized by the CPU, the memory, the program loaded in the memory, etc. in terms of hardware components, but in this case the functional blocks realized by the cooperation of them are depicted. Therefore, it is understood by those skilled in the art that these functional blocks can be realized in various forms by hardware only, software only, or a combination thereof.
- the determination result receiving unit 5 receives the final determination result by the inspector and the result of the nondestructive test of the object corresponding to the determination result from the plurality of plants 3.
- the learning unit 6 learns the determination algorithm 8 based on the information received by the determination result receiving unit 5.
- the providing unit 7 provides the determination device 30 of the plurality of plants 3 with the determination algorithm 8 learned by the learning unit 6.
- the learning device 4 is shown as a single device in the figure for simplification of the explanation, the learning device 4 is realized by a plurality of servers using cloud computing technology, distributed processing technology, etc. May be As a result, a large amount of information collected from a plurality of plants 3 can be processed at high speed to make the judgment algorithm 8 learn, so the time required to improve the accuracy of the judgment algorithm 8 can be significantly shortened. .
- FIG. 2 shows the configuration of the determination apparatus according to the embodiment.
- the determination apparatus 30 includes a test result acquisition unit 31, a determination unit 32, a determination result presentation unit 33, a final determination result acquisition unit 34, a determination result transmission unit 35, a learning unit 36, an update unit 37, and a determination algorithm 38. These configurations can also be realized in various forms by hardware only, software only, or a combination thereof.
- the local data server 40 stores a test result database 41, an AI determination result database 42, and a final determination result database 43.
- the test result acquisition unit 31 acquires the result of the nondestructive test of the inspection object 10.
- the nondestructive test results of the inspection object 10 by the test apparatus 20 are stored in the test result database 41.
- radiographs taken and developed by the test apparatus 20 are stored in the test result database 41.
- the test result acquisition unit 31 reads the test result from the test result database 41.
- the determination unit 32 determines pass / fail according to the determination algorithm 38 based on the nondestructive test result of the inspection object 10 acquired by the test result acquisition unit 31.
- the judgment algorithm 38 may detect various flaws and defects such as poor penetration, poor fusion, blow holes, pipes, slag inclusions, cracks, tungsten inclusions, etc., which may occur in the welds.
- the characteristic image pattern has been learned, and the determination unit 32 detects these characteristic image patterns present in the image, and determines the pass / fail of the detected flaw type or size in light of the test standard. judge.
- the determination unit 32 stores the determination result in the AI determination result database 42.
- the determination result presentation unit 33 reads the determination result by the determination unit 32 from the AI determination result database 42, and presents the result on the display device of the inspector terminal 39 used by the inspector who carries out the nondestructive inspection of the object.
- nondestructive inspection such as radiological transmission inspection (RT) that determines the soundness of welding using X-ray images of welds
- RT radiological transmission inspection
- pass / fail judgment is automatically made prior to final pass / fail judgment by visual inspection of inspectors. It is possible to improve the efficiency and accuracy of the nondestructive inspection because it is possible to sort out those that are suspected of being rejected.
- the automatic preliminary determination by the determination unit 32 only the image of the gray zone where the pass / fail determination is difficult may be determined by the inspector, so the number of man-hours of the inspector can be significantly reduced. As a result, the burden on individual inspectors can be greatly reduced, and labor costs can be reduced. Moreover, since the dispersion
- the final determination result acquisition unit 34 acquires, from the inspector terminal 39, a final determination result by the inspector who has confirmed the determination result by the determination unit 32.
- the determination result transmission unit 35 transmits the final determination result acquired by the final determination result acquisition unit 34 to the learning device 4 together with the nondestructive test result of the inspection object 10 corresponding to the determination result.
- the final determination result acquiring unit 34 may further acquire a comment by the inspector from the inspector terminal 39, and the determination result transmitting unit 35 may further transmit a comment by the inspector to the learning device 4.
- the comment by the inspector may be used to learn the determination algorithm 8 in the learning device 4. Thereby, the accuracy of the determination algorithm 8 can be further improved.
- the test results of the nondestructive testing performed in a plurality of plants 3 and the final determination results determined by the qualified inspector are collected in the learning device 4 to learn the determination algorithm 8
- learning data for learning the judgment algorithm 8 can be instantaneously taken into the learning device 4 without being influenced by the time or place where the nondestructive inspection is performed, and used for learning of the judgment algorithm 8.
- the speed of improving the accuracy of the determination algorithm 8 can be accelerated.
- the learning efficiency and learning speed of the judgment algorithm Can be significantly improved, and the accuracy of the determination algorithm can be accelerated in an accelerated manner.
- the type and pattern of flaws that are likely to occur may differ, but the judgment algorithm is unique in each plant 3.
- the algorithm for determining the flaws frequently generated in the plant 3 learns to a high accuracy
- the algorithm for determining the flaws hardly occurring in the plant 3 is the learning that can not be determined It may not progress.
- the inspection support system 1 of the present embodiment it is possible to collect information from a plurality of plants 3 and learn the judgment algorithm, so that various kinds of flaws can be appropriately detected to judge pass / fail. Possible accurate determination algorithms can be generated in a short period of time.
- the determination result transmission unit 35 acquires the corrected judgment result, and the result of the nondestructive test of the inspection object 10 corresponding to the judgment result. Together with the learning device 4. Thereby, the incorrect algorithm of the determination algorithm 8 can be corrected to improve the accuracy.
- the determination result transmission unit 35 acquires the determination result by the inspector with respect to the result of the nondestructive test of the inspection object 10 for which the determination unit 32 could not determine the pass / fail, and the inspection result of the inspection object 10 corresponding to the determination result.
- the update unit 37 acquires the determination algorithm 8 learned by the learning device 4 from the learning device 4 at a predetermined timing, and updates the determination algorithm 38.
- the nondestructive inspection can be performed more efficiently by using the determination algorithm 8 with improved accuracy.
- the determination device 30 can determine only the case where the determination is easy, and even if the determination by the inspector is performed for many cases, a sufficient amount of As information is collected and learning of the determination algorithm 8 progresses, the determination apparatus 30 can accurately determine in many cases, and the inspector only needs to perform a simple final confirmation.
- the learning unit 36 learns the determination algorithm 38 based on the final determination result by the inspector obtained by the final determination result acquisition unit 34 and the result of the nondestructive test of the object corresponding to the determination result. .
- the judgment algorithm 38 is subjected to reinforcement learning for cases that are likely to occur in each plant 3, and suitable for each plant 3 A high accuracy decision algorithm 38 can be generated.
- a radiation penetration test was mainly described as an example of nondestructive inspection.
- the nondestructive inspection to which the present invention can be applied is not limited to the radiation transmission test, for example, ultrasonic testing (UT), eddy current testing (ET), magnetic particle testing (Magnetic testing) Also applicable to particle testing (MT), penetration testing (PT), stress measurement (SM), acoustic emission (AE), thermographic test (IRT), etc. is there.
- the present invention is applicable to an inspection support system, a learning device, and a determination device for supporting nondestructive inspection of an object.
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Artificial Intelligence (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Signal Processing (AREA)
- Mathematical Physics (AREA)
- Computing Systems (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Molecular Biology (AREA)
- Quality & Reliability (AREA)
- Bioinformatics & Computational Biology (AREA)
- Medical Informatics (AREA)
- Toxicology (AREA)
- Multimedia (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Analysing Materials By The Use Of Radiation (AREA)
- Feedback Control In General (AREA)
- Alarm Systems (AREA)
Abstract
Description
Claims (9)
- 対象物の非破壊検査を支援するための検査支援システムであって、
前記対象物の非破壊試験の結果に基づいて合否を判定する複数の判定装置と、
前記複数の判定装置から収集される情報に基づいて、前記複数の判定装置において合否の判定に使用される判定アルゴリズムを学習させるための学習装置と、
を備え、
前記判定装置は、
前記対象物の非破壊試験の結果を取得する試験結果取得部と、
前記試験結果取得部により取得された前記対象物の非破壊試験の結果に基づいて、前記判定アルゴリズムにより合否を判定する判定部と、
前記判定部による判定結果を、前記対象物の非破壊検査を実施する検査員に提示する判定結果提示部と、
前記判定部による判定結果を確認した検査員による最終的な判定結果を取得し、その判定結果に対応する対象物の非破壊試験の結果とともに前記学習装置に送信する判定結果送信部と、
を備え、
前記学習装置は、
前記複数の判定装置から、検査員による最終的な判定結果と、その判定結果に対応する対象物の非破壊試験の結果を受信する判定結果受信部と、
前記判定結果受信部により受信された情報に基づいて、前記判定アルゴリズムを学習させる学習部と、
前記学習部により学習された前記判定アルゴリズムを、前記複数の判定装置に提供する提供部と、
を備えることを特徴とする検査支援システム。 - 前記判定結果送信部は、前記判定部による判定結果が検査員により修正された場合に、修正された判定結果を取得し、その判定結果に対応する対象物の非破壊試験の結果とともに前記学習装置に送信することを特徴とする請求項1に記載の検査支援システム。
- 前記判定結果送信部は、前記判定部により合否を判定できなかった前記対象物の非破壊試験の結果に対する検査員による判定結果を取得し、その判定結果に対応する対象物の非破壊試験の結果とともに前記学習装置に送信することを特徴とする請求項1又は2に記載の検査支援システム。
- 前記判定装置は、検査員による最終的な判定結果と、その判定結果に対応する対象物の非破壊試験の結果に基づいて、前記判定アルゴリズムを学習させる学習部を更に備えることを特徴とする請求項1から3のいずれかに記載の検査支援システム。
- 対象物の非破壊試験の結果に基づいて合否を判定する複数の判定装置から、前記判定装置による判定結果を確認した検査員による最終的な判定結果と、その判定結果に対応する対象物の非破壊試験の結果を受信する判定結果受信部と、
前記判定結果受信部により受信された情報に基づいて、前記複数の判定装置において合否の判定に使用される判定アルゴリズムを学習させる学習部と、
前記学習部により学習された前記判定アルゴリズムを、前記複数の判定装置に提供する提供部と、
を備えることを特徴とする学習装置。 - 前記判定結果受信部は、前記判定装置による判定結果が検査員により修正された場合に、修正された判定結果と、その判定結果に対応する対象物の非破壊試験の結果を受信することを特徴とする請求項5に記載の学習装置。
- 前記判定結果受信部は、前記判定装置により合否を判定できなかった前記対象物の非破壊試験の結果に対する検査員による判定結果と、その判定結果に対応する対象物の非破壊試験の結果を受信することを特徴とする請求項5又は6に記載の学習装置。
- 対象物の非破壊試験の結果を取得する試験結果取得部と、
前記試験結果取得部により取得された前記対象物の非破壊試験の結果に基づいて、合否の判定に使用される判定アルゴリズムを学習させるための学習装置から提供された前記判定アルゴリズムにより合否を判定する判定部と、
前記判定部による判定結果を、前記対象物の非破壊検査を実施する検査員に提示する判定結果提示部と、
前記判定部による判定結果を確認した検査員による最終的な判定結果を取得し、その判定結果に対応する対象物の非破壊試験の結果とともに前記学習装置に送信する判定結果送信部と、
を備えることを特徴とする判定装置。 - 検査員による最終的な判定結果と、その判定結果に対応する対象物の非破壊試験の結果に基づいて、前記判定アルゴリズムを学習させる学習部を更に備えることを特徴とする請求項8に記載の判定装置。
Priority Applications (4)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
AU2018371375A AU2018371375B2 (en) | 2017-11-21 | 2018-11-12 | Inspection assistance system, learning device, and assessment device |
EP18880256.5A EP3671188A4 (en) | 2017-11-21 | 2018-11-12 | INSPECTION ASSISTANCE SYSTEM, LEARNING DEVICE AND ASSESSMENT DEVICE |
CA3075911A CA3075911A1 (en) | 2017-11-21 | 2018-11-12 | Inspection assistance system, learning device, and assessment device |
US16/825,230 US11301976B2 (en) | 2017-11-21 | 2020-03-20 | Inspection support system, learning device, and determination device |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2017-223312 | 2017-11-21 | ||
JP2017223312A JP7050470B2 (ja) | 2017-11-21 | 2017-11-21 | 検査支援システム、学習装置、及び判定装置 |
Related Child Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US16/825,230 Continuation US11301976B2 (en) | 2017-11-21 | 2020-03-20 | Inspection support system, learning device, and determination device |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2019102892A1 true WO2019102892A1 (ja) | 2019-05-31 |
Family
ID=66630982
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2018/041847 WO2019102892A1 (ja) | 2017-11-21 | 2018-11-12 | 検査支援システム、学習装置、及び判定装置 |
Country Status (6)
Country | Link |
---|---|
US (1) | US11301976B2 (ja) |
EP (1) | EP3671188A4 (ja) |
JP (1) | JP7050470B2 (ja) |
AU (1) | AU2018371375B2 (ja) |
CA (1) | CA3075911A1 (ja) |
WO (1) | WO2019102892A1 (ja) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112179691A (zh) * | 2020-09-04 | 2021-01-05 | 西安交通大学 | 基于对抗学习策略的机械装备运行状态异常检测***和方法 |
JP2021001774A (ja) * | 2019-06-20 | 2021-01-07 | 株式会社小松製作所 | 非破壊検査システム、学習済みの排ガス処理フィルタ検査モデルの製造方法、および学習用データの生成方法 |
Families Citing this family (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2020079780A1 (ja) * | 2018-10-17 | 2020-04-23 | 株式会社島津製作所 | 航空機検査支援装置および航空機検査支援方法 |
WO2021033792A1 (ko) * | 2019-08-19 | 2021-02-25 | 엘지전자 주식회사 | Ai 기반의 사전 훈련 모델 결정 시스템 및 이를 이용한 제품 생산 라인을 위한 ai 기반의 비전 검사 관리 시스템 |
KR102377628B1 (ko) * | 2019-11-07 | 2022-03-24 | 한국전자통신연구원 | 인공지능 서비스에 대한 성능 관리 장치 및 방법 |
JP6890193B1 (ja) * | 2020-03-30 | 2021-06-18 | 東邦瓦斯株式会社 | 継手判別方法および継手判別装置 |
US11301980B2 (en) * | 2020-04-21 | 2022-04-12 | GM Global Technology Operations LLC | System and method to evaluate the integrity of spot welds |
KR102393109B1 (ko) * | 2020-04-29 | 2022-05-03 | 한국전력공사 | 이종 기관 간 협업을 위한 빅데이터 플랫폼 및 전역 학습 모델을 이용한 빅데이터 학습 방법 |
JP7137592B2 (ja) * | 2020-05-16 | 2022-09-14 | 三菱電機株式会社 | 劣化診断システム、劣化診断装置および劣化診断プログラム |
WO2022014019A1 (ja) * | 2020-07-16 | 2022-01-20 | 日揮グローバル株式会社 | 検査システム、判定処理装置、及び検査方法 |
JP7476082B2 (ja) | 2020-11-12 | 2024-04-30 | 株式会社東芝 | 非破壊検査の支援装置、支援方法及び支援プログラム |
KR102390004B1 (ko) * | 2021-09-16 | 2022-04-25 | 라이트브라더스 주식회사 | 스케일 변화량 분석 기반의 자전거 비파괴 검사 장치, 방법, 및 컴퓨터 판독 가능한 기록 매체 |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH0518904A (ja) | 1991-07-09 | 1993-01-26 | Nippon Steel Corp | 鋼管の溶接部検出方法 |
JP2013192624A (ja) * | 2012-03-16 | 2013-09-30 | Hitachi Ltd | 医用画像診断支援装置、医用画像診断支援方法ならびにコンピュータプログラム |
JP2014070944A (ja) * | 2012-09-28 | 2014-04-21 | Dainippon Screen Mfg Co Ltd | 教師データ検証装置、教師データ作成装置、画像分類装置、教師データ検証方法、教師データ作成方法および画像分類方法 |
WO2017073373A1 (ja) * | 2015-10-30 | 2017-05-04 | 株式会社モルフォ | 学習システム、学習装置、学習方法、学習プログラム、教師データ作成装置、教師データ作成方法、教師データ作成プログラム、端末装置及び閾値変更装置 |
WO2017081984A1 (ja) * | 2015-11-11 | 2017-05-18 | 株式会社日立製作所 | 制御装置及び診断システム |
JP2017142739A (ja) * | 2016-02-12 | 2017-08-17 | キヤノン株式会社 | 情報処理装置、情報処理方法及びプログラム |
Family Cites Families (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH10148622A (ja) * | 1996-11-18 | 1998-06-02 | Kawasaki Steel Corp | 溶接部放射線透過試験画像の自動合否判定方法及び装置 |
JP3730040B2 (ja) | 1998-12-14 | 2005-12-21 | Jfe工建株式会社 | X線検査装置及びx線検査方法 |
GB2362711B (en) * | 2000-05-25 | 2004-09-08 | Spectral Fusion Technologies L | Analysis of samples |
JP2004340667A (ja) | 2003-05-14 | 2004-12-02 | Kawasaki Heavy Ind Ltd | 可搬式x線撮影装置 |
US20050117700A1 (en) * | 2003-08-08 | 2005-06-02 | Peschmann Kristian R. | Methods and systems for the rapid detection of concealed objects |
JP5211810B2 (ja) | 2008-04-03 | 2013-06-12 | 新日鐵住金株式会社 | 疵学習装置、疵学習方法、及びコンピュータプログラム |
JP2010169232A (ja) | 2009-01-26 | 2010-08-05 | Chugoku Electric Power Co Inc:The | 配管の亀裂箇所の点検修理方法 |
US8761490B2 (en) * | 2011-11-03 | 2014-06-24 | United Technologies Corporation | System and method for automated borescope inspection user interface |
EP3329433A1 (en) * | 2015-07-29 | 2018-06-06 | Illinois Tool Works Inc. | System and method to facilitate welding software as a service |
JP6696278B2 (ja) * | 2016-04-12 | 2020-05-20 | 日本製鉄株式会社 | ドリフト検査装置 |
-
2017
- 2017-11-21 JP JP2017223312A patent/JP7050470B2/ja active Active
-
2018
- 2018-11-12 WO PCT/JP2018/041847 patent/WO2019102892A1/ja unknown
- 2018-11-12 EP EP18880256.5A patent/EP3671188A4/en active Pending
- 2018-11-12 AU AU2018371375A patent/AU2018371375B2/en active Active
- 2018-11-12 CA CA3075911A patent/CA3075911A1/en active Pending
-
2020
- 2020-03-20 US US16/825,230 patent/US11301976B2/en active Active
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH0518904A (ja) | 1991-07-09 | 1993-01-26 | Nippon Steel Corp | 鋼管の溶接部検出方法 |
JP2013192624A (ja) * | 2012-03-16 | 2013-09-30 | Hitachi Ltd | 医用画像診断支援装置、医用画像診断支援方法ならびにコンピュータプログラム |
JP2014070944A (ja) * | 2012-09-28 | 2014-04-21 | Dainippon Screen Mfg Co Ltd | 教師データ検証装置、教師データ作成装置、画像分類装置、教師データ検証方法、教師データ作成方法および画像分類方法 |
WO2017073373A1 (ja) * | 2015-10-30 | 2017-05-04 | 株式会社モルフォ | 学習システム、学習装置、学習方法、学習プログラム、教師データ作成装置、教師データ作成方法、教師データ作成プログラム、端末装置及び閾値変更装置 |
WO2017081984A1 (ja) * | 2015-11-11 | 2017-05-18 | 株式会社日立製作所 | 制御装置及び診断システム |
JP2017142739A (ja) * | 2016-02-12 | 2017-08-17 | キヤノン株式会社 | 情報処理装置、情報処理方法及びプログラム |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2021001774A (ja) * | 2019-06-20 | 2021-01-07 | 株式会社小松製作所 | 非破壊検査システム、学習済みの排ガス処理フィルタ検査モデルの製造方法、および学習用データの生成方法 |
CN112198230A (zh) * | 2019-06-20 | 2021-01-08 | 株式会社小松制作所 | 非破坏检查*** |
JP7394545B2 (ja) | 2019-06-20 | 2023-12-08 | 株式会社小松製作所 | 非破壊検査システム、学習済みの排ガス処理フィルタ検査モデルの製造方法、および学習用データの生成方法 |
CN112179691A (zh) * | 2020-09-04 | 2021-01-05 | 西安交通大学 | 基于对抗学习策略的机械装备运行状态异常检测***和方法 |
CN112179691B (zh) * | 2020-09-04 | 2021-07-13 | 西安交通大学 | 基于对抗学习策略的机械装备运行状态异常检测***和方法 |
Also Published As
Publication number | Publication date |
---|---|
US20200234425A1 (en) | 2020-07-23 |
JP7050470B2 (ja) | 2022-04-08 |
EP3671188A4 (en) | 2021-05-19 |
CA3075911A1 (en) | 2019-05-31 |
JP2019095247A (ja) | 2019-06-20 |
EP3671188A1 (en) | 2020-06-24 |
AU2018371375A1 (en) | 2020-04-02 |
US11301976B2 (en) | 2022-04-12 |
AU2018371375B2 (en) | 2023-11-16 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
WO2019102892A1 (ja) | 検査支援システム、学習装置、及び判定装置 | |
Reber et al. | A new generation of ultrasonic in-line inspection tools for detecting, sizing and locating metal loss and cracks in transmission pipelines | |
CN115836218A (zh) | 检查装置、检查方法、以及检查程序 | |
US20130104659A1 (en) | Method for Non-Destructive Testing | |
JP3853751B2 (ja) | 非破壊検査システム | |
KR20220071462A (ko) | 클라우드 기반의 배관 비파괴 검사시스템 | |
JP4728762B2 (ja) | 超音波探傷画像処理装置 | |
KR101315391B1 (ko) | 전자기 유도 센서를 활용한 비파괴 용접품질 전수검사 시스템 및 그 전수검사 방법 | |
Susarev et al. | Hardware and software for diagnosis of joint welds of industrial pipelines | |
WO2022014019A1 (ja) | 検査システム、判定処理装置、及び検査方法 | |
US20180196013A1 (en) | Tracking defects with permanently installed sensors | |
KR20160013750A (ko) | 전자기 유도 센서를 활용한 비파괴 피로 검사 장치 및 그 검사 방법 | |
JP2020173168A (ja) | ライニングの非破壊劣化検査方法 | |
RU2815195C1 (ru) | Способ автоматизированного рентгеновского контроля изделий сложной формы и устройство для его осуществления | |
EP4246140A1 (en) | Software-based assistance for non-destructive inspection | |
JP7455046B2 (ja) | 水中検査装置および水中検査方法 | |
RU2243566C1 (ru) | Способ определения целесообразного числа неразрушающих контролей изделия | |
Kania et al. | Investigation and Assessment of Low-Frequency ERW Seam Imperfections by EMAT and CMFL ILI | |
Huang | 1. General and reviews | |
KR20160113069A (ko) | 전자기 유도 센서를 활용한 비파괴 피로 검사 장치 및 그 검사 방법 | |
JP2009198249A (ja) | 超音波検査データ評価装置及び超音波検査データ評価方法 | |
Shahriari et al. | Determination of flaw type and location using an expert module in ultrasonic nondestructive testing for weld inspection | |
Babu et al. | Productivity & Reliability Study of Magnetic Particle Testing & Eddy current Testing for Inspection of Construction Welds | |
RU2243542C2 (ru) | Тест-образец для радиационного контроля | |
CN118130487A (zh) | 一种基于脚手架的焊接检测方法及*** |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 18880256 Country of ref document: EP Kind code of ref document: A1 |
|
ENP | Entry into the national phase |
Ref document number: 3075911 Country of ref document: CA |
|
ENP | Entry into the national phase |
Ref document number: 2018880256 Country of ref document: EP Effective date: 20200317 |
|
ENP | Entry into the national phase |
Ref document number: 2018371375 Country of ref document: AU Date of ref document: 20181112 Kind code of ref document: A |
|
NENP | Non-entry into the national phase |
Ref country code: DE |