WO2022210499A1 - 機械の状態検出装置 - Google Patents
機械の状態検出装置 Download PDFInfo
- Publication number
- WO2022210499A1 WO2022210499A1 PCT/JP2022/014895 JP2022014895W WO2022210499A1 WO 2022210499 A1 WO2022210499 A1 WO 2022210499A1 JP 2022014895 W JP2022014895 W JP 2022014895W WO 2022210499 A1 WO2022210499 A1 WO 2022210499A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- classification
- processing unit
- damage
- machine
- state
- Prior art date
Links
- 238000012545 processing Methods 0.000 claims abstract description 454
- 238000001514 detection method Methods 0.000 claims abstract description 144
- 239000013598 vector Substances 0.000 claims abstract description 114
- 230000001186 cumulative effect Effects 0.000 claims description 55
- 230000008859 change Effects 0.000 claims description 20
- 238000000605 extraction Methods 0.000 claims description 19
- 239000000284 extract Substances 0.000 claims 2
- 238000000034 method Methods 0.000 description 169
- 230000008569 process Effects 0.000 description 125
- 230000006870 function Effects 0.000 description 56
- 238000013145 classification model Methods 0.000 description 47
- 238000010586 diagram Methods 0.000 description 40
- 230000003252 repetitive effect Effects 0.000 description 12
- 230000009471 action Effects 0.000 description 11
- 238000005452 bending Methods 0.000 description 9
- 230000005856 abnormality Effects 0.000 description 6
- 238000012800 visualization Methods 0.000 description 6
- 238000010276 construction Methods 0.000 description 4
- 230000003247 decreasing effect Effects 0.000 description 4
- 238000012423 maintenance Methods 0.000 description 4
- 238000005259 measurement Methods 0.000 description 4
- 239000004576 sand Substances 0.000 description 4
- 239000003086 colorant Substances 0.000 description 3
- 238000005520 cutting process Methods 0.000 description 3
- 239000002689 soil Substances 0.000 description 3
- 238000007599 discharging Methods 0.000 description 2
- 238000009826 distribution Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 125000004122 cyclic group Chemical group 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000015654 memory Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000003909 pattern recognition Methods 0.000 description 1
- 230000008439 repair process Effects 0.000 description 1
- 238000010079 rubber tapping Methods 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
- 238000004148 unit process Methods 0.000 description 1
- 238000003466 welding Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B23/00—Testing or monitoring of control systems or parts thereof
- G05B23/02—Electric testing or monitoring
- G05B23/0205—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults
- G05B23/0259—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults characterized by the response to fault detection
- G05B23/0275—Fault isolation and identification, e.g. classify fault; estimate cause or root of failure
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M5/00—Investigating the elasticity of structures, e.g. deflection of bridges or air-craft wings
- G01M5/0033—Investigating the elasticity of structures, e.g. deflection of bridges or air-craft wings by determining damage, crack or wear
-
- 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
- G01M99/008—Subject matter not provided for in other groups of this subclass by doing functionality tests
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B23/00—Testing or monitoring of control systems or parts thereof
- G05B23/02—Electric testing or monitoring
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B23/00—Testing or monitoring of control systems or parts thereof
- G05B23/02—Electric testing or monitoring
- G05B23/0205—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults
- G05B23/0218—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults characterised by the fault detection method dealing with either existing or incipient faults
- G05B23/0224—Process history based detection method, e.g. whereby history implies the availability of large amounts of data
- G05B23/024—Quantitative history assessment, e.g. mathematical relationships between available data; Functions therefor; Principal component analysis [PCA]; Partial least square [PLS]; Statistical classifiers, e.g. Bayesian networks, linear regression or correlation analysis; Neural networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/20—Administration of product repair or maintenance
Definitions
- the present invention relates to a device for detecting the state of machines.
- Patent Literature 1 describes a device that learns the operating tendency of a machine when an abnormality occurs in the machine.
- Patent Literature 2 describes a detection device that detects a section on a time axis in which stress applied to a machine occurs, or a section at a position where the machine exists.
- Patent Literature 3 describes a damage estimating device that estimates the amount of damage received by a machine based on recognition results of actions performed by the machine.
- JP 2018-97616 A Japanese Patent No. 6215446 Japanese Patent No. 6458055
- Patent Documents 1 to 3 above have the following problems.
- (a) Japanese Patent Laid-Open No. 2002-200000 does not describe a means for determining whether the operation of the machine when an abnormality occurs is a known operation or an unknown operation, and an output method.
- the operating tendency of the machine is learned using alarms that indicate machine abnormalities, etc., it is also described about detecting the damage that the machine received (the cause of the abnormality) before the alarm that indicates an abnormality is generated. do not have.
- Patent Document 2 a section in which damage (repeated stress) occurs before an abnormality (machine breakage) occurs is detected, and the cause of the damage is specified using positional state recognition means.
- the positional state recognizing means used here recognizes a preset state, and there is no description of what kind of processing is to be performed when an unset state occurs.
- a table is created in advance that associates a motion recognition result with the amount of damage that a machine receives when performing that motion.
- the motion performed by the machine is recognized by a motion recognition method, and based on the recognition result, the damage amount corresponding to the motion is referenced to estimate the damage amount.
- the action recognition method used here is the recognition of a preset action as in (b) above, and what kind of processing is performed when an action that is not set is performed will be described. Not done.
- the present invention has been made in view of the above, and the state of the machine can be easily detected regardless of whether or not an unknown operation that is not assumed to cause damage has been performed. It is an object of the present invention to provide a detection device.
- a machine condition detection apparatus includes a collection device for collecting, at a plurality of times, condition information indicating a machine condition that changes in chronological order; and an arithmetic processing unit that processes a plurality of pieces of the state information to detect the state of the machine, wherein the arithmetic processing unit processes the plurality of pieces of state information collected by the collecting device.
- a classification processing unit that assigns, to a feature vector, a classification ID that identifies the set to which the feature vector belongs; a detection processing unit that detects damage received by the machine, a section in which the damage is detected by the detection processing unit, and the classification ID given by the classification processing unit in the section based on and a processing unit.
- the state of the machine can be easily detected regardless of whether or not an unknown action that is not supposed to cause damage has been performed. Problems, configurations, and effects other than those described above will be clarified by the following description of the embodiments.
- FIG. 2 is a processing block diagram of the state detection device according to the first embodiment
- FIG. 4 is a processing flow chart of the state detection device
- Explanatory drawing of repetitive stress amplitude and wavelength An example of information recorded in a classification model database.
- An example of state classification sensor/control information An example of the processing result of the classification process.
- Explanatory drawing of the pattern of classification ID An example of information recorded in the damage state classification ID database (No. 1). Another example of cutting out the damaged section.
- Part 3 An example of information recorded in the damage state classification ID database
- FIG. 10 is a processing block diagram of the state detection device according to the second embodiment; Explanatory drawing of cluster distribution of classification ID.
- FIG. 10 is a processing block diagram of the state detection device according to the second embodiment; Explanatory drawing of cluster distribution of classification ID.
- FIG. 11 is a processing block diagram of the state detection device of Embodiment 3; An example of information recorded in the damage state classification ID database (No. 4). Explanatory diagram of key frame animation. An example of information recorded in the damage state classification ID database (No. 5).
- FIG. 11 is a processing block diagram of the state detection device of Embodiment 4;
- FIG. 11 is a processing block diagram of the state detection device of Embodiment 5;
- Example of cumulative damage Example of chronological changes in cumulative damage.
- FIG. 11 is an example of visualization of an integrated value of individual damage amounts performed in the state detection device of Embodiment 6;
- FIG. An example visualization of the number of occurrences of the damage state.
- An example of a display screen in the state detection device of Embodiment 7 Part 1).
- FIG. 1 is a processing block diagram of the state detection device 100 of the first embodiment.
- FIG. 2 is a processing flow diagram of the state detection device 100.
- FIG. 3 is an explanatory diagram of repetitive stress amplitude and wavelength.
- FIG. 4 is a diagram showing an example of information recorded in the classification model database 1h.
- FIG. 5 is a diagram showing an example of state classification sensor/control information 1f.
- FIG. 6 is a diagram showing an example of the processing result of the classification processing 1i.
- FIG. 7 is an explanatory diagram of classification ID patterns.
- FIG. 8 is a diagram showing an example (part 1) of information recorded in the damaged state classification ID database 1o.
- the state detection device 100 is a device that detects the state of various machines, including construction machines and working machines such as industrial robots.
- the state detection device 100 is a device that detects the state of the machine that changes in time series (for example, the operating state of the machine).
- 1c, 1d, and 1e of FIG. 1 show an example of a machine to which the state detection device 100 is applied.
- 1c is a transport vehicle such as a dump truck.
- 1d is a vehicle-type construction machine such as a shovel.
- 1e is an industrial robot used in a manufacturing factory or the like.
- the state detection device 100 is not limited to these, and can be applied to various machines.
- the state detection device 100 is roughly divided into a collection device 1a that collects information for detecting the state of the machine, and an arithmetic processing device 1b that performs processing for detecting the state of the machine using the collected information. be able to. That is, the state detection device 100 includes a collection device 1a that collects state information indicating the state of the machine that changes in time series at a plurality of times, and a state information that is collected at a plurality of times and processed to process the state information of the machine. and an arithmetic processing unit 1b for detecting the state of
- the collection device 1a is composed of, for example, a device that stores various sensor information provided in the machine, and is composed of a controller or the like mounted on the machine.
- Each of the collection device 1a and the arithmetic processing device 1b may be configured by a general-purpose or industrial computer, or may be configured by dedicated hardware using a microcomputer.
- Each of the collection device 1a and the arithmetic processing device 1b includes a CPM, a ROM, and a RAM, and realizes their functions by the CPU executing programs stored in the ROM.
- the arithmetic processing unit 1b may transfer the information collected by the collecting device 1a to an external cloud computer using a communication line, and process the information using the cloud computer.
- the collection device 1a and the processing device 1b are shown separately, but the state detection device 100 accommodates the collection device 1a and the processing device 1b in one housing. It may be configured as a device.
- the collecting device 1a includes a state classification sensor/control information collection unit 1f that collects state classification sensor/control information 1f as state information indicating the state of the machine that changes in time series, and a damage detection sensor/control information collection unit 1f. and a damage detection sensor/control information collecting unit 1g for collecting 1g.
- the state classification sensor/control information 1f is sensor measurement information and information for controlling each part of the machine, which are used to classify the state of the machine.
- the damage detection sensor/control information 1g is sensor measurement information and information for controlling each part of the machine, which are used to detect damage received by the machine.
- the state classification sensor/control information 1f and the damage detection sensor/control information 1g may be information of the same type or information of different types.
- the state classification sensor/control information 1f is also referred to as "state information 1f".
- the damage detection sensor/control information 1g is also referred to as "state information 1g".
- the collection device 1a directly acquires the sensor measurement information included in the state information 1f and the state information 1g from a strain gauge or an angle sensor attached to the machine, or from a CAN (Controller Area Network) or the like mounted on the machine. It can be collected by acquiring information flowing on the network of The collecting device 1a can collect the control information included in the state information 1f and the state information 1g by directly obtaining it from a machine control device or the like, or by obtaining information flowing through a network mounted on the machine. can.
- Arithmetic processing unit 1b includes a detection processing unit 1j that executes detection processing 1j for damage received by the machine, a determination processing unit 1k that executes determination processing 1k for the amount of damage received by the machine, and the data collected by collection device 1a. Based on the state information 1f, a classification processing unit 1i that executes a classification processing 1i for classifying the state of the machine at each time, a processing result of the detection processing 1j (and the determination processing 1k), and a processing result of the classification processing 1i.
- a tying processing unit 1l that executes the tying process 1l, a determination processing unit 1m that executes a determination process 1m for determining whether the classification ID linked by the tying process 1l is an unknown classification ID, and a tying process 1l.
- a display processing unit 1n that executes display processing 1n for displaying that the assigned category ID is an unknown category ID; and a registration processing unit 1p for executing a registration processing 1p for registration.
- the arithmetic processing unit 1b includes a damaged state classification ID database 1o and a classification model database 1h as recording devices.
- the configuration of each unit in the collection device 1a and the processing unit 1b and the processing in each unit are represented by the same reference numerals.
- the arithmetic processing unit 1b has a function (detection processing unit 1j) for executing the detection processing 1j for damage received by the machine, as described above.
- the detection processing unit 1j performs processing for detecting damage received by the machine based on the state information 1g collected by the collection device 1a.
- the detection process 1j detects a section (period) on the time axis in which the damage received by the machine is large.
- the damage targeted by the detection process 1j is, for example, damage caused by stress fatigue damage to each part of a machine such as a hydraulic cylinder or a hydraulic motor.
- the damage targeted by the detection process 1j includes not only physical damage such as stress fatigue damage, but also safety problems such as mechanical imbalance. Machine imbalance conditions can be detected by attaching an inclinometer to the machine.
- FIG. 3 shows an example in which detection processing 1j is performed using this method.
- the horizontal axis in FIG. 3 indicates time.
- the vertical axis in FIG. 3 indicates stress (3a), stress amplitude (3b), and fatigue damage rate (3c) together with their respective time axes.
- 3h of FIG. 3 is a stress waveform, graphed from measurements of strain gauges mounted at predetermined locations on the machine. Fatigue damage depends on the amplitude of repeated stress rather than the absolute value of stress. 3h in FIG. 3 shows the absolute value of the stress, and it is not possible to know where the large fatigue damage occurred.
- the amplitude and wavelength of the repetitive stress are detected from the stress waveform 3h, and the section where the fatigue damage is large is detected.
- This method is a method of detecting the wavelength of stress by improving the rainflow method, which is a kind of stress cycle counting method.
- 3i and 3j in FIG. 3 are the amplitude and wavelength of repetitive stress detected using this method from the stress waveform of 3h.
- Rectangles 3i and 3j indicate the amplitude and wavelength of the detected stress.
- the length along the vertical axis of the rectangles 3i and 3j indicates the amplitude of repeated stress.
- the length along the horizontal axis of the rectangles 3i and 3j indicates the wavelength of the repetitive stress.
- the times corresponding to the vertices on the left side of the bases of the rectangles 3i and 3j indicate the start times of repeated stress.
- the time corresponding to the vertex on the right side of the base of the rectangles 3i and 3j indicates the end time of the repeated stress.
- the operation of loading earth and sand onto a dump truck is as follows: the dump truck moves in an empty state (3d), the earth and sand are loaded into the stopped dump truck in multiple stages (3e), and the dump truck moves in a loaded state. (3f), and (3g) to discharge the loaded earth and sand.
- 3j in FIG. 3 shows the amplitude and wavelength of the repetitive stress that the dump truck receives when the dump truck is loaded with earth and sand (3e).
- 3i of FIG. 3 shows the amplitude and wavelength of the repetitive stress that the dump truck receives when the dump truck is discharging soil (3g).
- the detection process 1j is based on the magnitude of damage received by the machine (magnitude of amplitude of repetitive stress) and the section in which the machine was damaged (the period from the start time to the end time of the amplitude of repetitive stress) and can be detected.
- 3c in FIG. 3 shows the fatigue damage (fatigue damage rate) that the machine receives per unit time.
- This fatigue damage rate 3c is obtained by calculating various fatigue damages received per unit time from various repeated stresses based on the amplitude and wavelength of the repeated stress shown in 3b, and integrating these fatigue damages per unit time. Calculated by Thereby, the detection process 1j can detect the magnitude of the damage at each time.
- Patent Document 2 is an example of a method for detecting the extent of damage received by a machine and the section in which the machine was damaged.
- Other techniques may be used for the detection process 1j as long as the magnitude of the damage received by the machine and the section in which the machine was damaged can be detected.
- the magnitude of damage received by the machine is also referred to as “amount of damage”.
- a section on the time axis in which the machine receives damage (a period from the start time of the damage to the end time of the damage) is also referred to as a “damage section”.
- the state of the machine when the machine receives damage (damage period) is also referred to as a "damaged state”.
- the amount of damage received by the machine in the damage section is also referred to as "individual damage amount”.
- the arithmetic processing unit 1b has a function (determination processing unit 1k) for executing the determination processing 1k of the amount of damage received by the machine, as described above.
- the determination processing unit 1k determines whether or not the amount of damage detected by the detection processing 1j is an amount of damage that affects the life of each part of the machine. In the damage amount determination process 1k, if the amount of damage detected by the detection process 1j is equal to or greater than a predetermined threshold value, it is determined that the damage amount affects the service life. Determined as the amount of damage that does not occur.
- the process of acquiring the state information 1g and the state information 1f is performed (corresponding to "NO" in 2a of FIG. 2).
- the later-described linking process 1l is performed (corresponding to "YES" in 2a of FIG. 2).
- the processing unit 1b may perform the linking process 1l for all damage amounts detected by the detection process 1j without setting a threshold in the determination process 1k. For example, if the damage amount is small but the damage state occurs frequently, it may be difficult to set a threshold for determining the damage amount that affects the life. In such a case, the arithmetic processing unit 1b may perform the linking process 1l for all damage amounts detected by the detection process 1j without performing the determination process 1k.
- the arithmetic processing unit 1b has a function (classification processing unit 1i) for executing the classification processing 1i for classifying the state of the machine at each time based on the state information 1f collected by the collection device 1a as described above.
- the classification processing unit 1i time-divides a plurality of pieces of state information 1f collected at a plurality of times by the collection device 1a, and performs a process of specifying a plurality of feature quantity vectors having the type of the state information 1f as a feature quantity. Then, the classification processing unit 1i classifies the specified feature amount vector into one of a plurality of predetermined sets (clusters), and for the classified feature amount vector, the set to which the feature amount vector belongs ( cluster) is assigned.
- the classification process 1i is performed based on information pre-recorded in the classification model database 1h.
- items (types) of the state information 1f are stored as feature amount vector items (vector components determined as feature amounts) (if there are n items of sensor/control information for state classification, n-dimensional feature
- a feature vector indicating the state and a classification ID for identifying the state are associated and recorded in advance.
- FIG. 4 shows an example of information recorded in the classification model database 1h.
- 4a in FIG. 4 shows the items of the feature amount vector from the sensor A to the operation amount Z.
- FIG. 4b in FIG. 4 shows the classification ID.
- the classification model database 1h can be constructed using, for example, a clustering method.
- the classification processing 1i a plurality of pieces of state information 1f for learning, which are collected at a plurality of times by the collection device 1a and record various states of the machine, are divided in time series to specify a plurality of feature amount vectors. do.
- the classification processing 1i classifies the plurality of specified feature amount vectors into one of a plurality of clusters using a clustering method, and assigns a classification ID.
- K-means is a method of grouping similar feature amount vectors into clusters of a number corresponding to the number of divisions set in advance, and returning a classification ID for identifying each cluster.
- the feature quantity vector divides the representative state of the machine by the set number of divisions m and indicates the characteristics of each state.
- the feature amount vectors to which "3" is assigned as the classification ID are ( M31, M32, ..., M3n ) .
- classification IDs There are m types of classification IDs, which are the number of divisions.
- the feature amount vectors recorded in the classification model database 1h shown in FIG. 4 are also referred to as "classification model feature amount vectors".
- a classification ID indicates one state (or scene) of a machine. For example, if the machine is a dump truck, the classification ID indicates a state in which the vessel of the dump truck is raised to remove soil, or if the machine is a shovel, the front arm is raised to excavate. Also, the state of the machine is indicated by a feature vector.
- the classification model database 1h associates and records a feature quantity vector indicating the state of the machine and a classification ID for identifying the feature quantity vector.
- the classification model database 1h of the first embodiment directly associates a natural language indicating the state of the machine, such as "the state in which the dump truck's vessel is raised and discharging soil" explained as an example, with the classification ID. It is not attached and recorded.
- the classification process 1i searches for the classification ID of the feature amount vector most similar to the feature amount vector of the state information 1f to be classified among the feature amount vectors recorded in the classification model database 1h. Being similar means that the distance (eg Euclidean distance) between the feature amount vectors is closer than others.
- the classification processing 1i searches for the feature quantity vector closest to the feature quantity vector to be classified among the feature quantity vectors recorded in the classification model database 1h, and assigns the classification ID of the searched feature quantity vector to the classification.
- the classification ID indicates the state of the machine at the time when the state information 1f on which the target feature amount vector was based was collected.
- FIG. 5 shows an example of the state information 1f on which the feature vectors to be classified are based.
- the state information 1f at “time T 3 ” in FIG. 5 is (F 13 , F 23 , . . . , F m3 ).
- This (F 13 , F 23 , . . . , F m3 ) constitutes a feature quantity vector whose feature quantity is each item (type) of the state information 1f at “time T 3 ”.
- the feature amount vector of the state information 1f shown in FIG. 5 is also referred to as "collected data feature amount vector".
- the classification process 1i searches for the feature amount vector of the classification model that is most similar to the feature amount vector of the collected data at each time shown in FIG. 5, and acquires the classification ID of the feature amount vector of the searched classification model.
- the degree of similarity between the feature amount vector of the collected data and the feature amount vector of the classification model is the distance between the feature amount vector of the collected data and the feature amount vector of the classification model (for example, Euclidean distance) can be calculated. Then, the classification process 1i may acquire the classification ID of the feature amount vector of the classification model with the closest distance.
- the Mahalanobis distance may be calculated in addition to the Euclidean distance, and various distance calculation methods can be used.
- a processing result of the classification processing 1i a sequence of classification IDs arranged in time series as shown in FIG. 6 is obtained.
- the arithmetic processing unit 1b has a function (association processing unit 1l) that executes the association processing 1l between the processing result of the detection processing 1j (and the determination processing 1k) and the processing result of the classification processing 1i.
- the linking processing unit 1l performs a process of linking a damage section, which is a section in which damage is detected by the damage detection processing 1j, and the classification ID given by the classification processing 1i in the damage section.
- the processing unit 1b can identify the damage state, which is the state of the machine in the damage section, by the classification ID assigned by the classification processing 1i.
- FIG. 7 shows a diagram in which the processing results of the detection processing 1j (and the determination processing 1k) and the processing results of the classification processing 1i are arranged in chronological order.
- 7a in FIG. 7 shows the processing result of the detection process 1j (and determination process 1k), and shows the amplitude and wavelength of the repetitive stress, like 3j in FIG. 7d in FIG. 7 indicates the starting time of the repeated stress, and 7e in FIG. 7 indicates the ending time of the repeated stress.
- 7b and 7c shown in the middle and lower stages of FIG. 7 indicate the processing result of the classification processing 1i.
- Parenthesized numbers such as “(3)” shown in the middle and lower stages of FIG. 7 indicate the value of the classification ID.
- the processing results of the classification process 1i are obtained as a column of classification IDs arranged in chronological order.
- the column of classification IDs in the section between the start time 7d and the end time 7e is composed of all classification IDs (5).
- Another type is a type in which the column of classification IDs has a characteristic pattern in the damage section between the start time 7d and the end time 7e, as shown in 7c of FIG. .
- the column of classification IDs in the damage section is configured with patterns of classification IDs (6), (6), (4), (7), (7), and (7). .
- a type in which the column of classification IDs consists of the same classification ID is observed when the machine moves less.
- a type in which the column of classification IDs is configured with a pattern is observed when the machine performs an action corresponding to the pattern. Since changes in the state of the machine appear as changes in the category ID, the column of category IDs will have a pattern corresponding to the state of the machine.
- the column of classification IDs can specify the damage state, which is the state of the machine during the damage interval.
- the arithmetic processing unit 1b has a function (determination processing unit 1m) that executes determination processing 1m for determining whether or not the classification ID linked by the linkage processing 1l is an unknown classification ID. Based on the classification IDs recorded in the damaged state classification ID database 1o, the determination processing unit 1m performs processing for determining whether or not the classification ID linked by the linkage processing 1l is an unknown classification ID. .
- the determination process 1m performs different processes depending on the type of the above-described classification ID string obtained as a result of the classification process 1i.
- the determination process 1m refers to the record of the category ID database 1o in which only one category ID is recorded if the column of the above category IDs is of a type composed of the same category ID.
- the determination process 1m is performed if the category ID (5) linked to the damaged section is recorded in the category ID database 1o, and the linked category ID is a known category ID. It is determined that there is (not an unknown classification ID).
- the determination processing 1m determines that the associated category ID is an unknown category ID.
- the determination process 1m refers to the record of the classification ID database 1o in which the items shown in FIG. 8 are recorded if the above classification ID string is of a type configured with a pattern.
- 8a in FIG. 8 shows the item of the pattern of the classification ID.
- 8b in FIG. 8 shows an ID item for identifying the pattern of the classification ID of 8a.
- Pattern recognition processing by DP matching may be used for this similarity determination.
- DP matching is a pattern matching method that uses dynamic programming, and is a method that considers elastic matching of patterns.
- Elastic matching means that, for example, in 7c of FIG. 7, the fact that classification IDs (6) and (7) appear in succession a plurality of times is considered to be "stretching", and the effect of "stretching" is eliminated.
- matching processing is performed assuming that classification IDs (6) and (7) appear once. Therefore, the pattern of the classification ID shown in 7c of FIG. 7 and the pattern of the classification ID shown in #1 of 8a in FIG. 8 are recognized as the same pattern. Note that DP matching can also calculate similarity.
- the similarity is 100%, and if only some of the category IDs included in the pattern match, the similarity is less than 100%.
- the determination process 1m if the similarity is not 100% but high to some extent, the pattern of the classification ID linked to the damaged section and the pattern of the classification ID recorded in the classification ID database 1o are determined to be the same pattern. You may In the determination process 1m, if the pattern of the classification ID associated with the damaged section is the same as the pattern of the classification ID recorded in the classification ID database 1o, the associated classification ID is a known classification ID. (It is not an unknown classification ID). On the other hand, in the determination process 1m, if the pattern of the category ID linked to the damaged section is not the same as the pattern of the category ID recorded in the category ID database 1o, the linked category ID is an unknown category ID. It is determined that
- the association processing 1l determines that the associated classification ID by the association processing 1l is a known classification ID. If it is determined that the associated classification ID by the association processing 1l is a known classification ID, the processing of acquiring the state information 1g and the state information 1f is performed (corresponding to "NO" in 2b of FIG. 2). . On the other hand, if the associated classification ID is determined to be an unknown classification ID by the association processing 1l, display processing 1n and registration processing 1p, which will be described later, are performed (corresponding to "YES" in 2b of FIG. 2). .
- the arithmetic processing unit 1b has a function (display processing unit 1n) that executes a display process 1n for displaying that the classification ID associated by the association process 1l is an unknown classification ID.
- the fact that the classification ID associated by the association process 1l is an unknown classification ID means that an unknown damage state has occurred.
- the display processing unit 1n performs processing for displaying information indicating that an unknown damage state has occurred on a display device such as a display, or for issuing a report using an audio output device such as a speaker. Note that the arithmetic processing unit 1b may omit the execution of the display processing 1n.
- the arithmetic processing unit 1b has a function (registration processing unit 1p) of executing a registration process 1p for registering the classification ID determined to be an unknown classification ID by the determination process 1m in the classification ID database 1o.
- the registration processing unit 1p performs processing to register the category ID (5) in the category ID database 1o.
- the registration processing unit 1p registers the pattern of the unknown classification ID in the part 8a described as "undefined" in #3 of FIG. 8b is registered.
- a classification ID that appears a plurality of times in a row is registered as having appeared once.
- the state detection device 100 of the first embodiment includes the collection device 1a that collects the state information 1f indicating the state of the machine that changes in chronological order at a plurality of times, and the state information 1f collected at a plurality of times. and an arithmetic processing unit 1b that processes the state information 1f and detects the state of the machine.
- the processing unit 1b time-divides the plurality of pieces of state information 1f collected by the collection device 1a, and classifies the feature quantity vectors having the type of the state information 1f as the feature quantity into one of a plurality of clusters (sets).
- a classification processing unit 1i for assigning a classification ID is provided.
- the arithmetic processing unit 1b includes a detection processing unit 1j that detects damage to the machine based on the state information 1g indicating the state of the machine that changes in time series and is collected by the collection unit 1a. Further, the arithmetic processing unit 1b includes an association processing unit 1l that associates a section in which damage is detected by the detection processing unit 1j with the classification ID given by the classification processing unit 1i in the section.
- the state detection device 100 of Embodiment 1 can identify the damage state, which is the state of the machine in the section in which the machine is damaged, by the classification ID given by the classification processing unit 1i. As a result, the state detection device 100 can easily detect that a damage state has occurred in the damage section linked to the classification ID, simply by referring to the classification ID assigned by the classification processing unit 1i. . Further, if the classification ID that identifies the damage state is recorded, the state detection device 100 detects that the detected damage state based on the recorded classification ID is an unknown state that is not assumed to cause damage. It can be easily determined whether or not there is. Therefore, the state detection device 100 of the first embodiment can easily detect the state of the machine regardless of whether or not an unknown action that is not assumed to cause damage has been performed.
- the processing device 1b includes a classification ID database 1o in which classification IDs identifying damage states are recorded in advance. Furthermore, the arithmetic processing unit 1b has a determination processing unit 1m that determines whether or not the classification ID linked by the linkage processing unit 1l is an unknown classification ID based on the classification ID recorded in the classification ID database 1o. Prepare. Further, the arithmetic processing unit 1b includes a registration processing unit 1p that registers, in the classification ID database 1o, a classification ID determined to be an unknown classification ID by the determination processing unit 1m.
- the state detection device 100 of the first embodiment can easily expand the classification ID database 1o in which the classification IDs specifying the damage state are recorded. It is possible to more reliably and easily determine whether or not there is. Since the state detection device 100 can reduce the frequency of occurrence of unknown damage states, it becomes easier to accurately detect the current state of the machine. Therefore, the state detection device 100 of the first embodiment can easily detect the state of the machine regardless of whether or not an unknown operation that is not supposed to cause damage has been performed, and can prevent maintenance of the machine. It is possible to make it easier to maintain the soundness of the machine, such as by facilitating timely execution.
- FIG. 9 is a diagram showing another example of cutting out a damaged section.
- FIG. 9 shows an example of cutting out a part of the fatigue damage rate waveform shown in 3c of FIG.
- a threshold value 9b that determines the strength of the waveform 9a is set, and the time 9c at which the waveform 9a changes to the threshold value 9b or more and the time 9d at which the waveform 9a changes to less than the threshold value 9b are set as the start time 9c and end time 9d of the damage section. can be detected.
- the detection process 1j may detect the maximum value of the waveform 9a in the damage section as the amount of damage received by the machine in the damage section. Alternatively, the detection process 1j may detect the average value of the waveform 9a in the damage section as the amount of damage received by the machine in the damage section.
- FIG. 10 is a diagram showing an example (part 2) of information recorded in the damaged state classification ID database 1o.
- FIG. 11 is a diagram showing an example (part 3) of information recorded in the damaged state classification ID database 1o.
- the linking process 1l not only links the damage section detected by the detection process 1j with the classification ID assigned by the classification process 1i in the damage section, but also the amount of damage received by the machine in the damage section ( Individual damage amount) may be linked. Then, as shown in FIGS. 10 and 11, in the category ID database 1o, the category ID assigned by the category processing 1i in the damage section may be recorded in association with the individual damage amount.
- FIG. 10 shows a case where the above-described column of classification IDs obtained as the processing result of the classification processing 1i is of a type composed of the same classification IDs.
- 10a in FIG. 10 indicates the category ID item.
- 10b in FIG. 10 shows the item of individual damage amount.
- the individual damage amount item 10b may be recorded separately for each part of the machine, as shown in FIG.
- FIG. 11 shows a case where the above-described classification ID string obtained as a processing result of the classification processing 1i is of a type configured with patterns.
- 11a in FIG. 11 shows the item of the pattern of the classification ID.
- 11b in FIG. 11 shows an ID item for identifying a pattern of classification IDs.
- 11c of FIG. 11 shows the item of individual damage amount.
- the individual damage amount item 11b may be recorded separately for each part of the machine, as in FIG. In the case of FIG. 11, the individual damage amount fluctuates within the damage section (see FIG. 9). It may be used as a representative value of the damage amount. Of course, a statistic other than the average value or the like may be used as the representative value of the individual damage amount in the damage section.
- the state detection device 100 of Embodiment 1 can identify not only the damage state but also the individual damage amount based on the classification ID assigned by the classification processing unit 1i. As a result, the state detection device 100 can not only easily detect that a damaged state has occurred, but also how much damage the machine has in the damaged section, simply by referring to the classification ID assigned by the classification processing unit 1i. It is possible to easily detect whether the amount of damage received is As a result, the state detection device 100 can more accurately detect the current fatigue damage degree of the machine. Therefore, the state detection device 100 of the first embodiment can easily and accurately detect the state of the machine, and can further maintain the soundness of the machine, such as performing maintenance of the machine in a more timely manner.
- the function of executing the detection processing 1j provided in the arithmetic processing unit 1b corresponds to an example of the "detection processing unit” described in the claims.
- the classification model database 1h provided in the arithmetic processing unit 1b corresponds to an example of the "second recording unit” described in the claims.
- the function of executing the classification processing 1i provided in the arithmetic processing unit 1b corresponds to an example of the "classification processing unit” described in the claims.
- the function of executing the determination processing 1m provided in the arithmetic processing device 1b corresponds to an example of the "first determination processing section" described in the claims.
- the damaged state classification ID database 1o provided in the processing unit 1b corresponds to an example of the "first recording unit” described in the claims.
- the function of executing the registration process 1p provided in the arithmetic processing unit 1b corresponds to an example of the "first registration processing unit” described in the claims.
- FIG. 12 is a processing block diagram of the state detection device 100 of the second embodiment.
- FIG. 13 is an explanatory diagram of the cluster distribution of classification IDs.
- the state detection device 100 of the first embodiment searches for the feature amount vector closest to the feature amount vector to be classified identified by the classification process 1i among the feature amount vectors pre-recorded in the classification model database 1h.
- the classification ID of the searched feature amount vector is used as the classification indicating the state of the machine at the time when the state information 1f on which the feature amount vector to be classified was collected. As an ID, it is assigned to the feature amount vector of the classification target.
- FIG. 13 shows a two-dimensional feature amount space in which one feature amount A of the feature amount vectors is described on the horizontal axis and another feature amount B of the feature amount vectors is described on the vertical axis. ing.
- the feature amount vector actually used in the present invention has a very large number of dimensions, the number of dimensions is reduced to simplify the explanation, and a two-dimensional feature amount space will be explained.
- Reference numerals 13a, 13b, 13c and 13d in FIG. 13 denote the position coordinates of the feature amount vectors of the classification IDs (5), (6), (1) and (3) pre-recorded in the classification model database 1h. show. 13e surrounded by 13g in FIG.
- the position coordinates 13g of the feature amount vector group are far from the position coordinates of the feature amount vectors pre-recorded in the classification model database 1h, and do not appear to belong to any classification ID cluster. This is probably because the learning interval for constructing the classification model database 1h did not include a machine state similar to the newly detected damage interval.
- the state detection device 100 of the second embodiment has processing functions as shown in FIG.
- the arithmetic processing unit 1b of the second embodiment performs determination processing for determining whether or not the classification ID of the feature vector similar to the feature vector to be classified identified by the classification processing 1i is recorded in the classification model database 1h. 12a (determination processing unit 12a). That is, the determination processing unit 12a determines the classification target specified by the classification process 1i based on the feature amount vector of the classification target specified by the classification process 1i and the feature amount vector recorded in advance in the classification model database 1h. A process of determining whether or not the classification ID to be assigned to the feature amount vector is recorded in the classification model database 1h is performed.
- the determination processing 12a calculates the distance between the feature amount vector to be classified and the feature amount vector recorded in the classification model database 1h. If the calculated distance is equal to or less than a preset threshold value, the determination processing 12a determines that the classification ID of the feature amount vector similar to the feature amount vector to be classified is recorded in the classification model database 1h. I judge. That is, in this case, the determination processing 12a determines that the classification ID to be assigned to the feature amount vector to be classified is recorded in the classification model database 1h. After that, the linking process 1l is performed in the same manner as in the first embodiment.
- the determination processing 12a if the calculated distance is larger (far) than a preset threshold value, the classification ID of the feature amount vector similar to the feature amount vector to be classified is recorded in the classification model database 1h. determine that it is not. That is, in this case, the determination processing 12a determines that the classification ID to be assigned to the feature amount vector to be classified is not recorded in the classification model database 1h. Thereafter, in the second embodiment, generation processing 12b and registration processing 12c, which will be described later, are performed.
- the processing unit 1b of the second embodiment executes the generation process 12b for generating a new classification model. It has a function (generation processing unit 12b). That is, the generation processing unit 12b performs processing for generating a new classification ID when the classification ID to be assigned to the feature amount vector to be classified is not recorded in the classification model database 1h.
- the arithmetic processing unit 1b of the second embodiment has a function (registration processing unit 12c) that executes a registration process 12c for registering the new classification model generated by the generation process 12b in the classification model database 1h. That is, the registration processing unit 12c performs a process of associating the new classification ID generated by the generation processing unit 12b with the feature vector to which the classification ID should be assigned and registering them in the classification model database 1h.
- registration processing unit 12c performs a process of associating the new classification ID generated by the generation processing unit 12b with the feature vector to which the classification ID should be assigned and registering them in the classification model database 1h.
- the generation process 12b and registration process 12c may be performed by either of the following two methods (method 1 or method 2).
- method 1 the generation processing 12b adds the feature amount vector group to be classified, which is specified in the damage interval newly detected this time, to the learning interval for constructing the classification model database 1h, and uses a classification method such as clustering. Then, a process of reconstructing the classification model database 1h is performed.
- the generation processing 12b performs the representative coordinate values (representative feature value of each feature quantity of the quantity vector). Then, the generation processing 12b generates a new classification ID to be assigned to the representative feature amount vector having the generated representative coordinate values.
- the registration process 12c associates the generated representative feature amount vector having the representative coordinate values with the generated new classification ID and registers them in the classification model database 1h.
- the registration process 12c registers the generated new category ID in the category ID database 1o as a category ID specifying the damage state.
- the generation process 12b calculates the representative coordinate values of the feature amount vector group from the group of feature amount vectors to be classified, which is specified in the damage section newly detected this time.
- the representative coordinate value of the feature vector group may be calculated, for example, by calculating the barycentric coordinate value or average coordinate value of the feature vector group.
- the generation process 12b generates a new classification ID to be assigned to the representative feature amount vector having the calculated representative coordinate values.
- the registration process 12c associates the calculated representative feature amount vector having the representative coordinate values with the generated new classification ID and registers them in the classification model database 1h.
- the registration process 12c registers the generated new category ID in the category ID database 1o as a category ID specifying the damage state.
- the state detection device 100 of the second embodiment can solve the above problems as described using FIG. 13 .
- Method 1 Normally, re-learning the classification model database 1h as in Method 1 imposes a large processing load on the processing unit 1b and takes a long time, so Method 2 is more effective.
- the generating process 12b performs method 2 for the first few times (the number of times is equal to or less than a preset threshold value), and performs method 2 when the newly generated classification IDs are accumulated. 1 may be used to re-learn the classification model database 1h.
- the state detection device 100 of the second embodiment generates a new classification ID when the classification ID to be assigned to the feature amount vector specified by the classification process 1i is not recorded, and the feature amount vector and the They are registered in the classification model database 1h in association with each other.
- the state detection device 100 of the second embodiment can easily expand the classification model database 1h in which the feature vector specifying the state of the machine is recorded, and can classify the state of the machine more reliably and easily. be able to. Since the state detection device 100 can reduce the frequency of recording the classification ID to be assigned to the feature amount vector specified by the classification processing 1i, the current state of the machine can be accurately detected. becomes easier. Therefore, the state detection device 100 of the second embodiment can easily detect the state of the machine regardless of whether or not an unknown operation that is not supposed to cause damage has been performed, and the maintenance of the machine can be easily performed. It is possible to make it easier to maintain the soundness of the machine, such as by facilitating timely execution.
- the function of executing the determination processing 12a provided in the arithmetic processing device 1b corresponds to an example of the "second determination processing section" described in the claims.
- the function of executing the generation processing 12b provided in the arithmetic processing device 1b corresponds to an example of the “generation processing unit” described in the claims.
- the function of executing the registration process 12c provided in the arithmetic processing unit 1b corresponds to an example of the "second registration processing unit" recited in the claims.
- FIG. 14 is a processing block diagram of the state detection device 100 of the third embodiment.
- FIG. 15 is a diagram showing an example (part 4) of information recorded in the damaged state classification ID database 1o.
- FIG. 16 is an explanatory diagram of keyframe animation.
- FIG. 17 is a diagram showing an example (No. 5) of information recorded in the damaged state classification ID database 1o.
- the arithmetic processing unit 1b of the third embodiment has a function of executing the extraction processing 14a for extracting the classification ID assigned by the classification processing 1i in a predetermined section including the section linked to the classification ID by the association processing 1l (extraction processing A portion 14a) is provided. That is, the extraction processing unit 14a performs processing for extracting the classification ID assigned by the classification processing 1i in the predetermined section.
- the classification ID extracted by the extraction process 14a is used when generating an animation in the reproduction process 14b, which will be described later.
- the predetermined section cut out by the cutout process 14a may be the damage section itself, which is the section linked to the classification ID by the linking process 1l, or may be a section including time before and after the damage section.
- the predetermined section cut out by the cutout processing 14a is a damage It is preferable that the section includes the time before and after the section (in the case of 7b in FIG. 7, the section includes the classification IDs (3) and (2) before and after the damage section). Note that even in this type, if it is sufficient to reproduce the state of the machine in the damaged section itself, the predetermined section cut out by the cutout processing 14a may be the damaged section itself. In this case, an input device that can be input by the user's operation may be provided to select the predetermined section to be cut out by the cutout processing 14a.
- the predetermined section cut out by the cutout processing 14a may be the damage section itself.
- the extraction processing 14a may integrate the category IDs that appear a plurality of times in succession, or may not integrate the category IDs that appear a plurality of times in succession. It may be left as it is.
- the animation is generated in the reproduction processing 14b, the animation reproduction speed is required.
- the classification IDs are integrated, it is preferable to attach time information such as the integration start time to the classification IDs.
- FIG. 15 shows an example in which items relating to information on the category ID for animation display are added to the category ID database 1o.
- 15a in FIG. 15 indicates an item related to the information of the classification ID for animation display.
- FIG. 15 shows an example in which classification IDs appearing consecutively are integrated.
- the arithmetic processing unit 1b of the third embodiment has a function (reproduction processing unit 14b) that executes a reproduction process 14b that generates an animation (moving image) that reproduces the state of the machine in the predetermined section from which the classification ID was extracted by the extraction process 14a.
- the reproduction processing unit 14b performs processing for generating an animation (moving image) that reproduces the state of the machine in a predetermined section based on the feature amount vector corresponding to the classification ID extracted by the extraction processing 14a.
- the reproduction process 14b uses keyframe animation used in computer graphics to generate an animation that reproduces the state of the machine in a predetermined section.
- FIG. 16 shows an example of generated animation that reproduces the state of a hydraulic excavator, which is one of the machines.
- the feature amount vector corresponding to the classification ID extracted by the extraction processing 14a includes information on the bending angle of each joint of the front working machine as the state of the machine (excavator). For example, information on bending angles of joints of booms, arms, and buckets.
- Keyframe animation is a method of generating this bending angle between classification IDs using interpolation. For example, it is assumed that the category IDs for animation display are category ID (11) indicated by 16f in FIG.
- the reproduction process 14b uses keyframe animation to generate frames that reproduce the state of the excavator between each of the frames 16a, 16b, and 16c so that the movement of the excavator can be smoothly reproduced. Perform frame interpolation.
- the reproduction process 14b performs keyframe interpolation for each bending angle of each joint of the shovel.
- 16i in FIG. 16 indicates the bending angle of the boom
- 16j indicates the bending angle of the arm
- 16k indicates the bending angle of the bucket.
- Each curve indicated by 16i, 16j and 16k in FIG. 16 represents the result of interpolating the bending angle of each joint between the three points 16f, 16g and 16h in FIG. 16 as the control points of the interpolation function. ing.
- interpolation function For example, a spline function or the like.
- the reproduction processing 14b can smoothly interpolate the bending angles of the joints at the times between the control points of the interpolation functions, and generate an animation that can smoothly reproduce the movement of the shovel.
- Figures 16d and 16e show the interpolated frames, interpolating between frames 16a and 16b and between frames 16b and 16c. Of course, it is also possible to generate frames that interpolate between frames 16a and 16d.
- the generation times of the classification IDs used as key frames are required.
- the reproduction processing 14b also interpolates the time information of the interpolated frame based on the time information of the key frame. Thereby, the reproduction processing 14b can reproduce the operating speed of the machine. Also, the reproduction process 14b may generate moving images other than animation.
- the arithmetic processing unit 1b of the third embodiment has a function (display processing unit) that executes display processing for displaying the animation (moving image) generated by the reproduction processing 14b on the display device.
- a function display processing unit
- the function of executing this display process may be included in the function of executing the display process 1n or the function of executing the reproduction process 14b. good.
- the state detection device 100 of the third embodiment cuts out the classification ID assigned by the classification process 1i in a predetermined section including the section linked to the classification ID by the linking process 1l. Then, the state detection device 100 generates an animation that smoothly reproduces the state of the machine based on the feature quantity vector corresponding to the extracted classification ID, and causes the display device to display the animation.
- the state detection device 100 of the third embodiment can visually express the state of the machine in the damaged section in an easy-to-understand manner, so that the user can intuitively grasp the damaged state. can. Also, in the state detection device 100 of the third embodiment, animation is generated using the feature amount vector corresponding to the classification ID, so animation can be generated with less information than the actually collected state information 1f.
- the state detection device 100 of the third embodiment does not generate an animation based on the feature amount vector corresponding to the extracted classification ID, that is, based on the state information 1f that constitutes the feature amount vector, but instead generates the animation based on the state information 1g.
- the state detection device 100 of the third embodiment may convert the content of the generated animation into text and register it in the classification ID database 1o as a text display character string as shown in 17a in FIG. At this time, the text display character string 17a is recorded in the category ID database 1o in association with the animation display category ID information 15a.
- the function of executing the extraction processing 14a provided in the arithmetic processing unit 1b corresponds to an example of the "extraction processing unit” described in the claims.
- the function of executing the reproduction processing 14b provided in the arithmetic processing unit 1b corresponds to an example of the "reproduction processing section” described in the claims.
- the function of executing the animation display process included in the function of executing the reproduction process 14b or the display process 1n provided in the arithmetic processing unit 1b corresponds to an example of the "display processing unit" described in the claims.
- FIG. 18 is a processing block diagram of the state detection device 100 of the fourth embodiment.
- the state detection device 100 of the fourth embodiment recognizes the damage state without using the state information 1g.
- a sensor retrofitted to the machine and a device for acquiring control information were required. Recognition is possible.
- the arithmetic processing unit 1b of the fourth embodiment has a function (recognition processing unit 18a) for executing the recognition processing 18a for recognizing the damage state specified by the classification ID assigned by the classification processing 1i.
- the recognition processing unit 18a recognizes the damage state specified by the classification ID given by the classification process 1i based on the classification ID given by the classification process 1i and the classification ID recorded in the classification ID database 1o. process.
- the recognition processing 18a searches for whether or not the above-described class ID string obtained as the processing result of the classifying processing 1i is recorded in the class ID database 1o as a class ID pattern as shown in 8a of FIG. 17, for example. process. If the column of the classification ID is of a type composed of the same classification ID (#3 in FIG. 17), the recognition processing 18a determines whether or not the classification ID is recorded in the classification ID database 1o and searches for it. Just do it. In the case of a type in which the row of the relevant classification ID has a pattern (#1 or #2 in FIG. 17), the recognition processing 18a searches using a pattern matching method such as the DP matching described above. Just do it. Thereby, the recognition processing 18a can detect the damage state without using the state information 1g.
- the time at which the first classification ID in the sequence of classification IDs obtained as the processing result of the classification processing 1i is assigned is the start time of the damaged state, and the time at which the last classification ID is assigned. is the end time of the damaged state, the damaged section can be recognized.
- the arithmetic processing unit 1b of the fourth embodiment has a function (display processing unit 18b) for executing the display processing 18b for displaying the recognition result of the damage state by the recognition processing 18a on the display device.
- the recognition processing 18a can search for the above-described classification ID string obtained as the processing result of the classification processing 1i being recorded in the classification ID database 1o
- the display processing unit 18b performs, for example, the animation display classification shown in FIG. It is possible to acquire the ID 15a and the text display character string 17a.
- the display processing unit 18b uses the acquired classification ID 15a for animation display and character string 17a for text display to perform processing for displaying an animation that reproduces the damaged state on the display device.
- the state detection device 100 of Embodiment 4 can recognize the damage state without using the state information 1g. As a result, even if the sensor or the like that acquires the state information 1g fails, the state detection device 100 of the fourth embodiment can easily detect the damage state based on the classification ID assigned by the classification processing unit 1i. .
- the state detection device 100 of the fourth embodiment can easily determine whether or not the detected damage state is an unknown state in which no damage is assumed, based on the classification ID recorded in the classification ID database 1o. can judge. Therefore, the state detection device 100 of the fourth embodiment can easily detect the state of the machine regardless of whether or not an unknown operation that is not assumed to cause damage has been performed.
- the function of executing the recognition processing 18a provided in the arithmetic processing device 1b corresponds to an example of the "recognition processing section" described in the claims.
- the function of executing the display processing 18b provided in the arithmetic processing device 1b corresponds to an example of the "display processing section" described in the claims.
- FIG. 19 is a processing block diagram of the state detection device 100 of the fifth embodiment.
- FIG. 20 is a diagram showing an example of cumulative damage degrees.
- FIG. 21 is a diagram showing an example of chronological changes in cumulative damage.
- the arithmetic processing device 1b has a function (estimation processing unit 19a) of performing a damage amount estimation process 19a based on the processing result of the recognition process 18a of the fourth embodiment.
- the classification ID database 1o as shown in FIGS. 10 and 11, the individual damage amount received by each part of the machine in the damage section is recorded in advance in association with the classification ID (or the pattern of the classification ID). good too.
- the estimation processing unit 19a recognizes the damage state from the recognition result of the recognition processing 18a, the estimation processing unit 19a performs processing for identifying the individual damage amount from the classification ID that identifies the damage state (for example, the classification ID (1) of #1 in FIG. 10). I do.
- the estimation processing unit 19a corresponds to the damage state recognized by the recognition processing 18a based on the category ID specifying the damage state recognized by the recognition processing 18a and the category ID recorded in the category ID database 1o. Perform processing to identify the individual damage amount to be applied. It should be noted that the individual damage amounts shown in FIGS. 10 and 11 are damage amounts that the machine receives due to one event.
- the estimation processing 19a can convert the individual amount of damage into a degree of damage using an SN diagram.
- the SN diagram shows the relationship between cyclic stress and the number of times the stress is repeated until the material breaks.
- the degree of damage is calculated from one repeated stress (stress amplitude) using an SN diagram.
- the cumulative damage level which is the cumulative value of the damage levels, reaches 1, it indicates that the material is fractured. In other words, the degree of damage can be said to be the consumption of the life of the machine. Therefore, the estimation processing 19a can convert the individual damage amounts shown in FIGS. 10 and 11 into damage degrees using the SN diagram.
- the estimation process 19a can obtain an SN diagram in advance by using the FAT class of the International Welding Society (IIW) or by actually performing a life test of the machine. If it is difficult to acquire the SN diagram, the estimation process 19a acquires the tendency of the degree of damage from the slope of a general SN diagram. In this case, the estimation processing 19a does not know the absolute value of the degree of damage, but can judge the tendency of the degree of damage (whether it is large or small). The estimation processing 19a calculates the degree of damage to estimate the influence of one damage state on the service life of the machine, or calculates the cumulative damage degree for each part of the machine to estimate the life consumption or life remaining amount. It is also possible to estimate the life of each part of the machine from the chronological change in the cumulative damage level.
- IIW International Welding Society
- the arithmetic processing unit 1b of the fifth embodiment has a function (display processing unit 19b) for executing display processing 19b for displaying the estimation result of the estimation processing 19a on the display device.
- the display processing unit 18b visualizes the influence that one damage state estimated by the estimation processing 19a has on the service life of the machine, visualizes the life consumption amount or the life remaining amount of each part of the machine, and Perform processing such as visualizing the prediction results of the lifespan of each part.
- FIG. 20 shows the cumulative degree of damage for each part of the machine.
- 20f in FIG. 20 indicates a line with a cumulative damage degree of 1.0.
- 20a-20e in FIG. 20 show the cumulative degree of damage of each part of the machine. It can be seen that the accumulated damage degree 20a of the site A in FIG. 20 is closest to the line of 20f and has the smallest life remaining amount.
- the display processing 19b may change the display mode of the cumulative damage level graph shown in FIG. 20 according to the remaining life amount. For example, the display processing 19b may change the color of the graph shown in FIG. 20 from green to yellow to red, etc., as the life remaining amount decreases.
- Fig. 21 shows chronological changes in the cumulative damage level.
- 21e in FIG. 21 indicates a line with a cumulative damage degree of 1.0.
- 21a to 21c in FIG. 21 show chronological changes in the degree of cumulative damage of each part of the machine.
- 21d in FIG. 21 is a line indicating the current time.
- 21f to 21h in FIG. 21 are straight lines showing prediction results of time-series changes 21a to 21c in the cumulative damage degree.
- the time indicated by the intersection of the straight lines 21f to 21h indicating the prediction result and the line 21e with the cumulative damage degree of 1.0 indicates the time when the service life is reached.
- the display processing 19b displays the remaining life time (the time from the time 21d to the time 21i), which is the time until the life reaches the end of the life, displays information prompting an order for parts for repair, or the like.
- the color of the graph shown in 21 can be changed and displayed in the same manner as in FIG. Further, the display processing 19b may display a drawing or the like of the portion when the graph shown in FIG. 20 or 21 is operated by the user (for example, clicked with a mouse or the like).
- the state detection device 100 of Embodiment 5 can estimate the individual damage amount from the processing result of the recognition processing 18a. Accordingly, the state detection device 100 of the fifth embodiment can calculate the degree of damage and the degree of cumulative damage from the estimated individual damage amount.
- the state detection device 100 of Embodiment 5 estimates the impact of a single damage state on the life of the machine, estimates the life consumption or life remaining for each part of the machine, and measures the cumulative degree of damage over time. It is also possible to predict the life of the machine from the change in temperature. Therefore, the state detection device 100 of the fifth embodiment can detect not only what state the machine was in during the damage period, but also how the state of the machine will change in the future from various viewpoints. can.
- the function of executing the estimation process 19a provided in the arithmetic processing device 1b of the fifth embodiment can also be provided in the arithmetic processing device 1b of the first to third embodiments.
- the arithmetic processing unit 1b of the first to third embodiments calculates the damage degree and the cumulative damage degree from the individual damage amount, estimates the life consumption amount or the life remaining amount for each part of the machine, or calculates the cumulative damage degree.
- a function for executing an estimation process 19a such as predicting the life of each part of the machine from time-series changes, may be provided.
- Patent Literature 3 described above it was possible to calculate the degree of damage only for a predetermined damage state, so it was not possible to estimate the degree of damage in an unexpected damage state. However, by providing the processing unit 1b of the first to third embodiments with the function of executing the estimation process 19a, it becomes possible to estimate the degree of damage in an unexpected damage state.
- the function of executing the estimation processing 19a provided in the arithmetic processing device 1b corresponds to an example of the "estimation processing unit” described in the claims.
- the function of executing the display processing 19b provided in the arithmetic processing unit 1b corresponds to an example of the "display processing section" described in the claims.
- FIG. 22 is a diagram showing an example of visualization of integrated values of individual damage amounts performed in the state detection device 100 of the sixth embodiment.
- FIG. 23 is a diagram illustrating an example of visualization of the number of occurrences of damage states.
- Embodiment 6 a method for identifying and visualizing the state of the machine that is the main cause of the amount of damage received by each part of the machine will be described.
- the arithmetic processing unit 1b of the fifth embodiment has functions for executing an estimation process 19a and a display process 19b. Damage and cumulative damage can be calculated and visualized.
- the arithmetic processing unit 1b of the sixth embodiment also has functions for executing the estimation processing 19a and the display processing 19b.
- the arithmetic processing unit 1b of the sixth embodiment can calculate and visualize the integrated value by accumulating the individual damage amount of a certain part recorded in the classification ID database 1o for each classification ID.
- the arithmetic processing device 1b of the sixth embodiment can estimate the cumulative damage amount by accumulating a plurality of calculated integrated values for each classification ID.
- the estimation processing 19a of the sixth embodiment performs the damage recognized by the recognition processing 18a based on the category ID specifying the damage state recognized by the recognition processing 18a and the category ID recorded in the category ID database 1o. Identify the individual damage amount corresponding to the state. Then, the estimation processing 19a of the sixth embodiment performs the individual damage amount based on the specified individual damage amount and the number of occurrences of the damage state corresponding to the individual damage amount (that is, the number of times the individual damage amount is received). The integrated value of the amount is calculated for each classification ID. Then, the estimation processing 19a of the sixth embodiment accumulates a plurality of integrated values calculated for each classification ID to estimate the cumulative amount of damage received by the machine. Then, the display processing 19b of the sixth embodiment causes the display device to display the processing result of the estimation processing 19a.
- the estimation processing 19a of the sixth embodiment multiplies the individual damage amount associated with a certain category ID by the number of occurrences of the damage state corresponding to the individual damage amount. An integrated value is calculated for each classification ID.
- the estimation processing 19a estimates the cumulative damage amount by calculating the sum of each of the multiple integrated values calculated for each category ID. Then, the estimation processing 19a of the sixth embodiment calculates the damage that is the main cause of the accumulated damage amount to the machine based on the ratio of each of the plurality of integrated values estimated for each classification ID to the accumulated damage amount. Identify the state.
- the damage state of category ID (1) occurs once and the damage state of category ID (3) occurs 1000 times.
- the ratio of the integrated value 5000 of the individual damage amount in the category ID(3) is larger than the ratio of the integrated value 1050 of the individual damage amount in the category ID(1). Therefore, it is specified that the damage state that is the main cause of the cumulative damage amount 6050 received by the part A is the damage state of classification ID (3). 22 and 23 visualize this situation.
- FIG. 22 shows the integrated value of the individual damage amount received by the part A for each classification ID.
- FIG. 23 shows the number of occurrences of the damage state corresponding to the individual damage amount received by the part A for each classification ID. From FIG. 22, it can be seen that the damage state that is the main cause of the cumulative amount of damage received by the site A is the damage state of classification ID (3). From FIG. 23, it can be seen that the damage state of classification ID (3) is a damage state with a large number of occurrences. can be reduced.
- the estimation processing 19a of the sixth embodiment can convert the cumulative damage amount into the cumulative damage degree, as in the fifth embodiment.
- the life consumption amount or the life remaining amount is estimated for each part of the machine, or from the time-series change of the cumulative damage degree life expectancy can be predicted.
- the processor 1b of the sixth embodiment combines the straight lines 21f to 21h showing the prediction results of the time series changes 21a to 21c of the cumulative damage degree as shown in FIG. 21 with the method of the present embodiment.
- a control command may be generated so as not to cause the damage state of category ID (3) so often, and sent to the control device that controls the operation of the machine.
- the classification ID (3) is the damage state classified by the clustering method, and it may be difficult for the user to understand what kind of state it is.
- the processing unit 1b of the sixth embodiment has a function of executing the extraction processing 14a and the reproduction processing 14b of the third embodiment, generates an animation that reproduces the damage state of the classification ID (3), and displays the can be displayed.
- the estimation processing 19a calculates the integrated value of the individual damage amount for each classification ID. , it is calculated for each ID that identifies the pattern of the classification ID. Further, in FIGS. 20, 22 and 23, the arithmetic processing unit 1b is visualized using a bar graph, but it may be visualized using other formats such as a pie chart.
- the integrated value of the individual damage amount is calculated for each classification ID, and a plurality of integrated values calculated for each classification ID are accumulated to determine the cumulative damage received by the machine. Estimate quantity.
- the state detection device 100 of the sixth embodiment from the ratio of each of the plurality of integrated values estimated for each classification ID to the cumulative damage amount, the damage that is the main cause of the cumulative damage amount to the machine is determined. state can be specified. Therefore, the state detection device 100 of the sixth embodiment can appropriately take measures such as controlling the machine so as not to cause the damage state that is the main cause, so that the soundness of the machine can be further maintained. .
- the function of executing the estimation processing 19a and the display processing 19b provided in the arithmetic processing device 1b of the sixth embodiment can also be provided in the arithmetic processing device 1b of the first to third embodiments.
- the arithmetic processing device 1b of the first to third embodiments based on the individual damage amount detected by the detection process 1j and the number of times the individual damage amount is detected (that is, the number of occurrences of the damage state), determines the individual damage amount is calculated for each classification ID, and a plurality of accumulated values calculated for each classification ID are accumulated to estimate the cumulative amount of damage received by the machine. As a result, even if an unexpected damage state is the main cause of the accumulated damage amount, the arithmetic processing device 1b of the first to third embodiments can appropriately estimate it.
- FIG. 24 is a diagram showing an example (part 1) of a display screen in the state detection device 100 of the seventh embodiment.
- FIG. 25 is a diagram showing an example (2) of the display screen.
- FIG. 26 is a diagram showing an example (part 3) of the display screen.
- the arithmetic processing unit 1b of the seventh embodiment has a function of executing the clipping process 14a and the reproduction process 14b, and also has a function of executing a display process for displaying the animation (moving image) generated by the reproduction process 14b on the display device. Prepare.
- the display processing of the seventh embodiment includes time-series changes in the status information 1f collected in the predetermined section from which the classification ID was cut out by the cutout processing 14a, time-series changes in the individual damage amount in the predetermined section, cumulative damage At least one of the amount, the chronological change in the accumulated damage amount, and the prediction result of the life of the machine is displayed on the display device together with the animation generated by the reproduction processing 14b.
- Fig. 24 shows an example of the display screen when the machine receives damage that affects its life.
- 24a in FIG. 24 shows the entire display screen.
- 24b in FIG. 24 shows the pressure waveform of the damaged bucket cylinder, that is, the chronological change of the state information 1f.
- 24i in FIG. 24 indicates the waveform of the individual damage amount detected based on the pressure waveform 24b of the bucket cylinder, that is, the chronological change in the individual damage amount.
- 24f in FIG. 24 shows an animation that reproduces the action (hereinafter also referred to as "damage action”) performed by the damaged machine in a predetermined section.
- the display device that displays the display screen 24a may be a display device installed in the driver's seat of the construction machine, or a display device installed in the machine's monitoring room. may be Display screens 25a and 26a, which will be described later, are also displayed on a similar display device.
- the display screen 24a may be displayed in real time immediately after the machine is damaged, or may be displayed at a predetermined timing such as after the machine is finished working.
- the display processing of the seventh embodiment may display the display screen 24a when the user operates a display button or the like.
- the display screens 25a and 26a are similarly displayed in real time or at a predetermined timing.
- the display processing of Embodiment 7 can display the display screen 24a on the display device when the machine receives damage.
- the display processing of the seventh embodiment includes information indicating that a damaged state has occurred, such as "a damaged state has occurred", and a Time information can be displayed to notify the user.
- the damaged machine part is displayed in text as shown in 24d in FIG. 24, or the color or shape of the damaged machine part is displayed as shown in 24h in FIG.
- the changed highlighting can be superimposed and displayed on the animation 24f.
- the highlighting may differ depending on the cumulative damage level. For example, if the cumulative damage level of the damaged machine part is high, it will be marked as “red”. In some cases, it may be displayed as "blue". Of course, the displayed colors are not limited to the three types of red, yellow, and blue, and these three types of colors are merely examples.
- the display processing of the seventh embodiment is not limited to highlighting the damaged parts of the machine.
- the animation 24f that displays the entire machine is displayed as the main parts of the machine (if the machine is an excavator, the boom , arms, buckets, traveling bodies, etc.) may be displayed in different colors.
- the display processing of the seventh embodiment can display the cumulative damage level of the entire machine, such as which part has a high cumulative damage level and which part requires attention, even for parts other than the highlighted parts. The user can easily understand each time he/she receives the
- the time-series change of the state information 1f and the time-series change of the individual damage amount are displayed to show what kind of machine the machine is. The user can be notified of how much damage has been received.
- FIG. 25 shows an example of the display screen when the life remaining amount of a certain part of the machine has decreased.
- 25a in FIG. 25 shows the entire display screen.
- 25b in FIG. 25 indicates the degree of cumulative damage of each part of the machine, that is, the cumulative amount of damage received by each part of the machine.
- 25f of FIG. 25 shows an animation similar to 24f of FIG.
- the display processing of Embodiment 7 can cause the display screen 25a to be displayed on the display device when the life remaining amount of a certain portion of the machine is reduced.
- the display processing of the seventh embodiment displays information indicating that the cumulative damage degree has increased and the life remaining amount has decreased as shown in 25d and 25e of FIG. As shown, it can be illustrated by a bar graph or the like.
- the display processing of the seventh embodiment can display the highlighted display of the part where the life remaining amount has decreased so as to be superimposed on the animation 25f.
- a damage state is associated with a classification ID, but may not be associated with a text display character string as indicated by 17a in FIG. Even in this case, the display processing of the seventh embodiment reproduces the damage state in which the remaining amount of life has decreased as the animation 25f, so that the user can be notified of the state of the machine in an easy-to-understand manner.
- FIG. 26 shows an example of a display screen that displays the result of predicting the service life of the machine.
- 26a in FIG. 26 shows the entire display screen.
- 26b in FIG. 26 shows the chronological change in the degree of cumulative damage in a certain part of the machine and the prediction result of its life. That is, 26b in FIG. 26 shows the chronological change in the cumulative damage amount in a certain part of the machine and the prediction result of its life.
- 26f of FIG. 26 shows an animation similar to 25f of FIG.
- the display processing of Embodiment 7 can predict the life of a certain portion of the machine when the life remaining amount is reduced, and display the display screen 26a on the display device.
- information indicating that the cumulative damage degree has increased and information indicating the remaining life time are displayed in text to notify the user. can do.
- the display processing of the seventh embodiment can notify the user by displaying the action to be taken by the user in text as shown in 26e of FIG. If the action to be taken by the user is to contact the agency, the user taps (or clicks after specifying with a pointer) the display 26e of "Contact the agency" to display the agency's phone number, etc. Contact information may be displayed, or procedures for simple countermeasures may be displayed.
- FIGS. 24 to 26 are examples of display screens, and the presence or absence of text display for notifying the user can be set arbitrarily.
- the display processing of the seventh embodiment may first display only the animations 24f to 26f. Further detailed information may be displayed by the user tapping the highlighted display superimposed on the animations 24f to 26f.
- a wide display area is secured for the animations 24f to 26f, allowing the user to accurately grasp which parts of the machine as a whole have been damaged (whether the degree of cumulative damage is high).
- only information that the user considers necessary can be notified.
- the state detection device 100 of the seventh embodiment can detect time-series changes in the sensor/control information 1f collected in a predetermined section including the damage section, time-series changes in the individual damage amount, cumulative damage amount, At least one of a chronological change in the accumulated damage amount and a prediction result of the life of the machine is displayed together with an animation that reproduces the damage state.
- the state detection device 100 of the seventh embodiment can visualize the damage state and also visualize changes in the state of the machine in the future from various viewpoints. can be done.
- the present invention is not limited to the above-described embodiments, and includes various modifications.
- the above embodiments have been described in detail in order to explain the present invention in an easy-to-understand manner, and are not necessarily limited to those having all the configurations described.
- it is possible to replace part of the configuration of one embodiment with the configuration of another embodiment and it is also possible to add the configuration of another embodiment to the configuration of one embodiment.
- each of the above configurations, functions, processing units, processing means, etc. may be realized by hardware, for example, by designing them in integrated circuits, in part or in whole.
- each of the above configurations, functions, etc. may be realized by software by a processor interpreting and executing a program for realizing each function.
- Information such as programs, tapes, and files that implement each function can be stored in recording devices such as memories, hard disks, SSDs (solid state drives), or recording media such as IC cards, SD cards, and DVDs.
- control lines and information lines indicate what is considered necessary for explanation, and not all control lines and information lines are necessarily indicated on the product. In practice, it may be considered that almost all configurations are interconnected.
Abstract
Description
(a)特許文献1では、異常が発生した時の機械の動作が、既知の動作なのか未知の動作なのかを判定する手段及び出力方法について述べられていない。また、機械の異常を示すアラーム等を用いて機械の動作傾向を学習しているので、異常を示すアラーム発生前に機械が受けたダメージ(異常の発生要因)を検出することに関しても述べられていない。
(b)特許文献2では、異常(機械の破断)を引き起こす前のダメージ(繰り返し応力)の発生した区間を検出し、ダメージを受けた要因を、位置状態認識手段を用いて特定している。ところが、ここで用いている位置状態認識手段は、予め設定された状態を認識するものであり、設定されていない状態が発生した場合にどのような処理を行うかに関しては述べられていない。
(c)特許文献3では、予め動作認識結果とその動作を行った場合に機械が受けるダメージ量とを紐付けるテーブルを作成する。そして、運用時には、機械の行った動作を動作認識手法により認識し、認識結果に基づき、当該動作に対応するダメージ量を参照して、ダメージ量を推定している。ところが、ここで用いている動作認識手法は、上記(b)と同様に予め設定された動作の認識であり、設定されていない動作が行われた場合にどのような処理を行うかに関しては述べられていない。
上記以外の課題、構成および効果は、以下の実施形態の説明により明らかにされる。
図1~図11を用いて、実施形態1の状態検出装置100について説明する。
図1は、実施形態1の状態検出装置100の処理ブロック図である。図2は、状態検出装置100の処理フロー図である。図3は、繰り返し応力振幅及び波長の説明図である。図4は、分類モデルデータベース1hに記録された情報の例を示す図である。図5は、状態分類用センサ・制御情報1fの例を示す図である。図6は、分類処理1iの処理結果の例を示す図である。図7は、分類IDのパターンの説明図である。図8は、ダメージ状態の分類IDデータベース1oに記録された情報の例(その1)を示す図である。
なお、説明の都合上、収集装置1a内及び演算処理装置1b内の各部の構成と、各部における処理とを同一の符号により表すものとする。
図12~図13を用いて、実施形態2の状態検出装置100について説明する。実施形態2の状態検出装置100において、従前の実施形態と同様の構成及び動作については、説明を省略する。
図12は、実施形態2の状態検出装置100の処理ブロック図である。図13は、分類IDのクラスタ分布の説明図である。
図14~図17を用いて、実施形態3の状態検出装置100について説明する。実施形態3の状態検出装置100において、従前の実施形態と同様の構成及び動作については、説明を省略する。
図14は、実施形態3の状態検出装置100の処理ブロック図である。図15は、ダメージ状態の分類IDデータベース1oに記録された情報の例(その4)を示す図である。図16は、キーフレームアニメーションの説明図である。図17は、ダメージ状態の分類IDデータベース1oに記録された情報の例(その5)を示す図である。
図18を用いて、実施形態4の状態検出装置100について説明する。実施形態4の状態検出装置100において、従前の実施形態と同様の構成及び動作については、説明を省略する。
図18は、実施形態4の状態検出装置100の処理ブロック図である。
図19~図21を用いて、実施形態5の状態検出装置100について説明する。実施形態5の状態検出装置100において、従前の実施形態と同様の構成及び動作については、説明を省略する。
図19は、実施形態5の状態検出装置100の処理ブロック図である。図20は、累積損傷度の例を示す図である。図21は、累積損傷度の時系列的な変化の例を示す図である。
図22~図23を用いて、実施形態6の状態検出装置100について説明する。実施形態6の状態検出装置100において、従前の実施形態と同様の構成及び動作については、説明を省略する。
図22は、実施形態6の状態検出装置100において行われる個別ダメージ量の積算値の可視化の例を示す図である。図23は、ダメージ状態の発生回数の可視化の例を示す図である。
図24~図26を用いて、実施形態7の状態検出装置100について説明する。実施形態7の状態検出装置100において、従前の実施形態と同様の構成及び動作については、説明を省略する。
図24は、実施形態7の状態検出装置100における表示画面の例(その1)を示す図である。図25は、表示画面の例(その2)を示す図である。図26は、表示画面の例(その3)を示す図である。
なお、本発明は上記の実施形態に限定されるものではなく、様々な変形例が含まれる。例えば、上記の実施形態は本発明を分かりやすく説明するために詳細に説明したものであり、必ずしも説明した全ての構成を備えるものに限定されるものではない。また、或る実施形態の構成の一部を他の実施形態の構成に置き換えることが可能であり、また、或る実施形態の構成に他の実施形態の構成を加えることも可能である。また、各実施形態の構成の一部について、他の構成の追加・削除・置換をすることが可能である。
Claims (9)
- 時系列的に変化する機械の状態を示す状態情報を複数の時刻において収集する収集装置と、前記複数の時刻において収集された複数の前記状態情報を処理して前記機械の状態を検出する演算処理装置と、を備える機械の状態検出装置であって、
前記演算処理装置は、
前記収集装置により収集された前記複数の状態情報を時分割して前記状態情報の種別を特徴量とする複数の特徴量ベクトルを特定し、特定された前記特徴量ベクトルを予め定められた複数の集合の何れかに分類し、分類された前記特徴量ベクトルに対して、当該特徴量ベクトルが属する前記集合を識別する分類IDを付与する分類処理部と、
前記収集装置により収集された時系列的に変化する前記機械の状態を示す状態情報に基づいて、前記機械が受けたダメージを検出する検出処理部と、
前記検出処理部により前記ダメージが検出された区間と、当該区間において前記分類処理部により付与された前記分類IDとを紐付ける紐付処理部と、を備える
ことを特徴とする機械の状態検出装置。 - 前記演算処理装置は、
前記機械がダメージを受けた前記区間における前記機械の状態であるダメージ状態を特定する前記分類IDが予め記録された第1記録部と、
前記第1記録部に記録された前記分類IDに基づいて、前記紐付処理部により紐付けられた前記分類IDが、未知の前記分類IDであるか否かを判定する第1判定処理部と、 前記第1判定処理部により前記未知の分類IDであると判定された前記分類IDを前記第1記録部に登録する第1登録処理部と、を更に備える
ことを特徴とする請求項1に記載の機械の状態検出装置。 - 前記演算処理装置は、
前記特徴量ベクトルと前記分類IDとが予め対応付けられて記録された第2記録部と、
前記分類処理部により特定された前記特徴量ベクトルと前記第2記録部に記録された前記特徴量ベクトルとに基づいて、前記分類処理部により特定された前記特徴量ベクトルに付与するべき前記分類IDが前記第2記録部に記録されているか否かを判定する第2判定処理部と、
前記分類処理部により特定された前記特徴量ベクトルに付与するべき前記分類IDが前記第2記録部に記録されていない場合、新たな前記分類IDを生成する生成処理部と、 前記生成処理部により生成された前記分類IDと、当該分類IDを付与するべき前記特徴量ベクトルとを対応付けて前記第2記録部に登録する第2登録処理部と、を更に備える
ことを特徴とする請求項1に記載の機械の状態検出装置。 - 前記演算処理装置は、
前記紐付処理部により前記分類IDに紐付けられた前記区間を含む所定区間において前記分類処理部により付与された前記分類IDを切り出す切出処理部と、
前記切出処理部により切り出された前記分類IDに基づいて、前記所定区間における前記機械の前記状態を再現する動画を生成する再現処理部と、
前記再現処理部により生成された前記動画を表示装置に表示させる表示処理部と、を更に備える
ことを特徴とする請求項1に記載の機械の状態検出装置。 - 前記演算処理装置は、前記機械が受けたダメージ量を推定する推定処理部を更に備え、 前記検出処理部は、前記区間において収集された前記状態情報に基づいて、前記区間において前記機械が受けたダメージ量である個別ダメージ量を検出し、
前記紐付処理部は、前記検出処理部により検出された前記個別ダメージ量を、当該個別ダメージ量が検出された前記区間において前記分類処理部により付与された前記分類IDに紐付け、
前記推定処理部は、
前記検出処理部により検出された前記個別ダメージ量と、当該個別ダメージ量の検出回数とに基づいて、当該個別ダメージ量の積算値を前記分類ID毎に算出し、
前記分類ID毎に算出された複数の前記積算値を累積して前記機械が受けた累積ダメージ量を推定する
ことを特徴とする請求項1に記載の機械の状態検出装置。 - 前記演算処理装置は、
前記紐付処理部により前記分類IDに紐付けられた前記区間を含む所定区間において前記分類処理部により付与された前記分類IDを切り出す切出処理部と、
前記切出処理部により切り出された前記分類IDに基づいて、前記所定区間における前記機械の前記状態を再現する動画を生成する再現処理部と、
前記再現処理部により生成された前記動画を表示装置に表示させる表示処理部と、を更に備え、
前記表示処理部は、前記所定区間において収集された前記状態情報の前記時系列的な変化、前記所定区間における前記個別ダメージ量の前記時系列的な変化、前記累積ダメージ量、前記累積ダメージ量の前記時系列的な変化、及び、前記機械の寿命の予測結果の少なくとも1つを、前記動画に併せて前記表示装置に表示させる
ことを特徴とする請求項5に記載の機械の状態検出装置。 - 時系列的に変化する機械の状態を示す状態情報を複数の時刻において収集する収集装置と、前記複数の時刻において収集された複数の前記状態情報を処理して前記機械の状態を検出する演算処理装置と、を備える機械の状態検出装置であって、
前記演算処理装置は、
前記収集装置により収集された前記複数の状態情報を時分割して前記状態情報の種別を特徴量とする複数の特徴量ベクトルを特定し、特定された前記特徴量ベクトルを予め定められた複数の集合の何れかに分類し、分類された前記特徴量ベクトルに対して、当該特徴量ベクトルが属する前記集合を識別する分類IDを付与する分類処理部と、
前記機械がダメージを受けた区間における前記機械の状態であるダメージ状態を特定する前記分類IDが予め記録された第1記録部と、
前記分類処理部により付与された前記分類IDと、前記第1記録部に記録された前記分類IDとに基づいて、前記分類処理部により付与された前記分類IDによって特定される前記ダメージ状態を認識する認識処理部と、を備える
ことを特徴とする機械の状態検出装置。 - 前記演算処理装置は、前記機械が受けたダメージ量を推定する推定処理部を更に備え、 前記第1記録部には、前記区間において前記機械が受けたダメージ量である個別ダメージ量が、前記分類IDに予め紐付けられて記録されており、
前記推定処理部は、
前記認識処理部により認識された前記ダメージ状態を特定する前記分類IDと、前記第1記録部に記録された前記分類IDとに基づいて、前記認識処理部により認識された前記ダメージ状態に対応する前記個別ダメージ量を特定し、
特定された前記個別ダメージ量と、当該個別ダメージ量に対応する前記ダメージ状態の発生回数とに基づいて、当該個別ダメージ量の積算値を前記分類ID毎に算出し、
前記分類ID毎に算出された複数の前記積算値を累積して前記機械が受けた累積ダメージ量を推定する
ことを特徴とする請求項7に記載の機械の状態検出装置。 - 前記演算処理装置は、
前記認識処理部により認識された前記ダメージ状態に対応する前記区間を含む所定区間において前記分類処理部により付与された前記分類IDを切り出す切出処理部と、
前記切出処理部により切り出された前記分類IDに基づいて、前記所定区間における前記機械の前記状態を再現する動画を生成する再現処理部と、
前記再現処理部により生成された前記動画を表示装置に表示させる表示処理部と、を更に備え、
前記表示処理部は、前記所定区間において収集された前記状態情報の前記時系列的な変化、前記所定区間における前記個別ダメージ量の前記時系列的な変化、前記累積ダメージ量、前記累積ダメージ量の前記時系列的な変化、及び、前記機械の寿命の予測結果の少なくとも1つを、前記動画に併せて前記表示装置に表示させる
ことを特徴とする請求項8に記載の機械の状態検出装置。
Priority Applications (4)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2023511237A JPWO2022210499A1 (ja) | 2021-04-02 | 2022-03-28 | |
CN202280016994.9A CN116981923A (zh) | 2021-04-02 | 2022-03-28 | 机械的状态检测装置 |
EP22780710.4A EP4317936A1 (en) | 2021-04-02 | 2022-03-28 | Machine state detecting device |
KR1020237029376A KR20230138504A (ko) | 2021-04-02 | 2022-03-28 | 기계의 상태 검출 장치 |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2021-063255 | 2021-04-01 | ||
JP2021063255 | 2021-04-02 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2022210499A1 true WO2022210499A1 (ja) | 2022-10-06 |
Family
ID=83456277
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2022/014895 WO2022210499A1 (ja) | 2021-04-02 | 2022-03-28 | 機械の状態検出装置 |
Country Status (5)
Country | Link |
---|---|
EP (1) | EP4317936A1 (ja) |
JP (1) | JPWO2022210499A1 (ja) |
KR (1) | KR20230138504A (ja) |
CN (1) | CN116981923A (ja) |
WO (1) | WO2022210499A1 (ja) |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH0158055B2 (ja) | 1981-04-23 | 1989-12-08 | Star Seiki Kk | |
WO2016117041A1 (ja) * | 2015-01-21 | 2016-07-28 | 株式会社日立製作所 | 損傷推定装置 |
JP6215446B2 (ja) | 2014-03-03 | 2017-10-18 | 株式会社日立製作所 | 機械の材料疲労の表示方法、及びその装置 |
WO2018042616A1 (ja) * | 2016-09-02 | 2018-03-08 | 株式会社日立製作所 | 診断装置、診断方法及び診断プログラム |
JP2018097616A (ja) | 2016-12-13 | 2018-06-21 | ファナック株式会社 | 数値制御装置及び機械学習装置 |
JP2018185256A (ja) * | 2017-04-27 | 2018-11-22 | 公益財団法人鉄道総合技術研究所 | 鉄道車両機器診断装置および鉄道車両機器診断方法 |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPS6215446U (ja) | 1985-07-12 | 1987-01-29 |
-
2022
- 2022-03-28 WO PCT/JP2022/014895 patent/WO2022210499A1/ja active Application Filing
- 2022-03-28 KR KR1020237029376A patent/KR20230138504A/ko unknown
- 2022-03-28 JP JP2023511237A patent/JPWO2022210499A1/ja active Pending
- 2022-03-28 EP EP22780710.4A patent/EP4317936A1/en active Pending
- 2022-03-28 CN CN202280016994.9A patent/CN116981923A/zh active Pending
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH0158055B2 (ja) | 1981-04-23 | 1989-12-08 | Star Seiki Kk | |
JP6215446B2 (ja) | 2014-03-03 | 2017-10-18 | 株式会社日立製作所 | 機械の材料疲労の表示方法、及びその装置 |
WO2016117041A1 (ja) * | 2015-01-21 | 2016-07-28 | 株式会社日立製作所 | 損傷推定装置 |
WO2018042616A1 (ja) * | 2016-09-02 | 2018-03-08 | 株式会社日立製作所 | 診断装置、診断方法及び診断プログラム |
JP2018097616A (ja) | 2016-12-13 | 2018-06-21 | ファナック株式会社 | 数値制御装置及び機械学習装置 |
JP2018185256A (ja) * | 2017-04-27 | 2018-11-22 | 公益財団法人鉄道総合技術研究所 | 鉄道車両機器診断装置および鉄道車両機器診断方法 |
Also Published As
Publication number | Publication date |
---|---|
JPWO2022210499A1 (ja) | 2022-10-06 |
KR20230138504A (ko) | 2023-10-05 |
CN116981923A (zh) | 2023-10-31 |
EP4317936A1 (en) | 2024-02-07 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
JP5945350B2 (ja) | 設備状態監視方法およびその装置 | |
JP5538597B2 (ja) | 異常検知方法及び異常検知システム | |
JP6060261B2 (ja) | 状態監視装置 | |
JP5179086B2 (ja) | 工業プロセスの監視方法及び監視システム | |
JP5431235B2 (ja) | 設備状態監視方法およびその装置 | |
WO2010095314A1 (ja) | 異常検知方法及び異常検知システム | |
JP6903142B2 (ja) | 異常検出装置、異常検出方法及びプログラム | |
JP5045770B2 (ja) | プロセス解析システム | |
JP5858839B2 (ja) | 作業機械の異常診断装置 | |
JP2021092970A (ja) | 制御方法、制御装置、機械設備、制御プログラム、記録媒体 | |
JPWO2020188696A1 (ja) | 異常検知装置および異常検知方法 | |
JP6475469B2 (ja) | 診断ジョブ生成システム、診断ジョブ生成方法及び診断ジョブ生成表示方法 | |
US20160098501A1 (en) | Virtual sensors supported by a computer aided design (cad) model and software | |
US20220108501A1 (en) | Information processing method, information processing apparatus, recording medium, method of manufacturing products, method of acquiring learning data, display method, and display apparatus | |
WO2022210499A1 (ja) | 機械の状態検出装置 | |
AU2005234089B2 (en) | System and method to query for machine conditions | |
JPH10124134A (ja) | 診断システム | |
JP2019159885A (ja) | 動作分析装置、動作分析方法、動作分析プログラム及び動作分析システム | |
US20240134367A1 (en) | Machine state detecting device | |
US20220088802A1 (en) | Information processing method, information processing apparatus, display method, display apparatus, recording medium, method of manufacturing products, and method of acquiring learning data | |
JP2021096639A (ja) | 制御方法、制御装置、機械設備、制御プログラム、記録媒体 | |
JP2020107248A (ja) | 異常判定装置および異常判定方法 | |
CN115616994A (zh) | 一种设备稳定性分析方法及系统、设备、存储介质 | |
JP7000266B2 (ja) | 分析フロー作成システムおよび分析フロー作成方法 | |
JP7193491B2 (ja) | 音点検システム |
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: 22780710 Country of ref document: EP Kind code of ref document: A1 |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2023511237 Country of ref document: JP |
|
WWE | Wipo information: entry into national phase |
Ref document number: 202280016994.9 Country of ref document: CN |
|
ENP | Entry into the national phase |
Ref document number: 20237029376 Country of ref document: KR Kind code of ref document: A |
|
WWE | Wipo information: entry into national phase |
Ref document number: 18279454 Country of ref document: US |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2022780710 Country of ref document: EP |
|
ENP | Entry into the national phase |
Ref document number: 2022780710 Country of ref document: EP Effective date: 20231102 |
|
NENP | Non-entry into the national phase |
Ref country code: DE |