WO2017061028A1 - 異常検知装置 - Google Patents
異常検知装置 Download PDFInfo
- Publication number
- WO2017061028A1 WO2017061028A1 PCT/JP2015/078752 JP2015078752W WO2017061028A1 WO 2017061028 A1 WO2017061028 A1 WO 2017061028A1 JP 2015078752 W JP2015078752 W JP 2015078752W WO 2017061028 A1 WO2017061028 A1 WO 2017061028A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- calculation result
- abnormality
- detection device
- abnormality detection
- data
- Prior art date
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/07—Responding to the occurrence of a fault, e.g. fault tolerance
- G06F11/0703—Error or fault processing not based on redundancy, i.e. by taking additional measures to deal with the error or fault not making use of redundancy in operation, in hardware, or in data representation
- G06F11/079—Root cause analysis, i.e. error or fault diagnosis
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B21—MECHANICAL METAL-WORKING WITHOUT ESSENTIALLY REMOVING MATERIAL; PUNCHING METAL
- B21B—ROLLING OF METAL
- B21B37/00—Control devices or methods specially adapted for metal-rolling mills or the work produced thereby
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B21—MECHANICAL METAL-WORKING WITHOUT ESSENTIALLY REMOVING MATERIAL; PUNCHING METAL
- B21B—ROLLING OF METAL
- B21B37/00—Control devices or methods specially adapted for metal-rolling mills or the work produced thereby
- B21B37/48—Tension control; Compression control
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B21—MECHANICAL METAL-WORKING WITHOUT ESSENTIALLY REMOVING MATERIAL; PUNCHING METAL
- B21B—ROLLING OF METAL
- B21B37/00—Control devices or methods specially adapted for metal-rolling mills or the work produced thereby
- B21B37/58—Roll-force control; Roll-gap control
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M15/00—Testing of engines
- G01M15/04—Testing internal-combustion engines
- G01M15/05—Testing internal-combustion engines by combined monitoring of two or more different engine parameters
-
- 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
- G05B23/0227—Qualitative history assessment, whereby the type of data acted upon, e.g. waveforms, images or patterns, is not relevant, e.g. rule based assessment; if-then decisions
- G05B23/0235—Qualitative history assessment, whereby the type of data acted upon, e.g. waveforms, images or patterns, is not relevant, e.g. rule based assessment; if-then decisions based on a comparison with predetermined threshold or range, e.g. "classical methods", carried out during normal operation; threshold adaptation or choice; when or how to compare with the threshold
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/07—Responding to the occurrence of a fault, e.g. fault tolerance
- G06F11/0703—Error or fault processing not based on redundancy, i.e. by taking additional measures to deal with the error or fault not making use of redundancy in operation, in hardware, or in data representation
- G06F11/0706—Error or fault processing not based on redundancy, i.e. by taking additional measures to deal with the error or fault not making use of redundancy in operation, in hardware, or in data representation the processing taking place on a specific hardware platform or in a specific software environment
- G06F11/0727—Error or fault processing not based on redundancy, i.e. by taking additional measures to deal with the error or fault not making use of redundancy in operation, in hardware, or in data representation the processing taking place on a specific hardware platform or in a specific software environment in a storage system, e.g. in a DASD or network based storage system
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/07—Responding to the occurrence of a fault, e.g. fault tolerance
- G06F11/0703—Error or fault processing not based on redundancy, i.e. by taking additional measures to deal with the error or fault not making use of redundancy in operation, in hardware, or in data representation
- G06F11/0751—Error or fault detection not based on redundancy
- G06F11/0754—Error or fault detection not based on redundancy by exceeding limits
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B21—MECHANICAL METAL-WORKING WITHOUT ESSENTIALLY REMOVING MATERIAL; PUNCHING METAL
- B21B—ROLLING OF METAL
- B21B2275/00—Mill drive parameters
- B21B2275/02—Speed
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B21—MECHANICAL METAL-WORKING WITHOUT ESSENTIALLY REMOVING MATERIAL; PUNCHING METAL
- B21B—ROLLING OF METAL
- B21B38/00—Methods or devices for measuring, detecting or monitoring specially adapted for metal-rolling mills, e.g. position detection, inspection of the product
- B21B38/06—Methods or devices for measuring, detecting or monitoring specially adapted for metal-rolling mills, e.g. position detection, inspection of the product for measuring tension or compression
-
- 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
- G05B2219/00—Program-control systems
- G05B2219/30—Nc systems
- G05B2219/36—Nc in input of data, input key till input tape
- G05B2219/36208—Roll grinding
Definitions
- the present invention relates to a technique for detecting an abnormality in a calculation result by software.
- Patent Document 1 states that “the time series data of the instrumentation system is used to improve the detection accuracy of a device that detects an abnormal sign such as a failure of a plant facility.
- the local time-series data extraction unit 102 inputs a plurality of sets of time-series data, and applies a predetermined time segmentation method to the sets to generate a plurality of time-series data.
- the local time-series data model estimation unit 103 uses the extracted local time-series data as a predetermined model estimation method.
- the local time-series data clustering unit 104 divides each model-estimated local time-series data into a plurality of clusters and obtains representative local parameters representing the clusters for each cluster. Then, the outlier detection unit 106 detects, based on the representative local parameter, whether the value defined in advance as a distance from any of the representative local parameters exceeds the threshold with respect to the evaluation target data for a predetermined period. To do. Is disclosed (see summary).
- the control software that implements the process for electronically controlling the controlled object is updated relatively frequently. This update is performed, for example, by distributing updated software via a network. Such frequently updated software may output an erroneous control command value due to an implementation error, for example.
- Patent Document 1 The technique described in Patent Document 1 is for detecting an abnormality of a controlled object such as a plant, particularly an outlier abnormality. Such an abnormality is mainly intended to detect an abnormality that has occurred in the controlled object. However, when an abnormality occurs in the control software itself, it is considered that more appropriate measures can be taken by detecting the abnormality of the control software rather than detecting the abnormality of the control target. The above-mentioned patent document 1 does not always sufficiently consider the detection of processing abnormality in the control software.
- the present invention has been made in view of the above-described problems, and an object thereof is to provide an abnormality detection device capable of detecting various software processing abnormalities.
- the anomaly detection device divides an output data series output by software into one or more clusters in advance, determines that output data included in any cluster is normal, and is included in any cluster. No output data is determined to be abnormal.
- the abnormality detection device can detect various processing abnormalities of software. Thereby, the reliability of the controlled object can be improved.
- FIG. 2 is a diagram illustrating a configuration of software stored in a ROM 120.
- FIG. 3 is a functional block diagram illustrating a detailed configuration of an abnormality detection unit 122.
- FIG. 10 is a diagram for explaining internal processing of a data divider 1221.
- FIG. 11 is a diagram illustrating an example of the result of clustering by a data divider 1221.
- FIG. 10 is a diagram for describing internal processing of a range setting unit 1222. It is a figure which shows the result by which the range setter 1222 set the normal range with respect to the clustering result shown in FIG. It is a figure explaining the internal process of the abnormality determination device 1223.
- FIG. 10 is a diagram for explaining internal processing of a range setting unit 1222. It is a figure which shows the result by which the range setter 1222 set the normal range with respect to the clustering result shown in FIG. It is a figure explaining the internal process of the abnormality determination device 1223.
- FIG. 10 is a diagram for explaining
- FIG. 6 is a diagram illustrating a configuration of software stored in a ROM 120 according to Embodiment 2.
- FIG. It is a functional block diagram which shows the detailed structure of the abnormality detection part 122 in Embodiment 2.
- FIG. It is a functional block diagram which shows the detailed structure of the abnormality detection part 122 in Embodiment 3.
- FIG. It is a figure which shows the example of the generation frequency of a time-dependent transition pattern which the time-dependent pattern frequency calculator 1224 calculated. It is a figure explaining the internal process of the abnormality determination device 1223 in Embodiment 3.
- FIG. FIG. 10 is a diagram illustrating a configuration of software stored in a ROM 120 according to a fourth embodiment.
- FIG. 10 is a functional block diagram illustrating a detailed configuration of an abnormality detection unit 122 according to a fifth embodiment. It is a figure which shows the specific example of the control object 200 in Embodiment 6. FIG. It is a figure which shows the specific example of the control object 200 in Embodiment 7.
- FIG. 20 is a functional block diagram illustrating a detailed configuration of an abnormality detection unit 122 according to an eighth embodiment. It is a figure which shows the abnormality detection apparatus 100 which concerns on Embodiment 9, and its periphery structure.
- FIG. 1 is a block diagram illustrating a configuration of an abnormality detection apparatus 100 according to Embodiment 1 of the present invention.
- the abnormality detection apparatus 100 itself detects an abnormality of software executed by the abnormality detection apparatus 100 will be described, but the present invention is not limited to this.
- the anomaly detection device 100 is a device for detecting an abnormality in a calculation result by software.
- the abnormality detection apparatus 100 includes a CPU (Central Processing Unit) 110, a ROM (Read Only Memory) 120, a RAM (Random Access Memory) 130, a data bus 140, an input circuit 150, an input / output port 160, and an output circuit 170.
- CPU Central Processing Unit
- ROM Read Only Memory
- RAM Random Access Memory
- the input circuit 150 processes an external signal.
- an external signal here, for example, a detection signal output from a sensor, a data signal from a server, a data signal from another control device or a data processing device, and the like can be considered. Signals from the outside are transmitted via a data bus, a dedicated line, radio, or the like.
- the input circuit 150 processes an external signal and then outputs it as an input signal to the input / output port 160.
- the input / output port 160 writes an input signal to the RAM 130 via the data bus 140.
- the ROM 120 stores software executed by the CPU 110.
- CPU 110 executes software using a value temporarily stored in RAM 130.
- the data value sent to the outside of the abnormality detection apparatus 100 is transmitted to the input / output port 160 via the data bus 140. This is output to the output circuit 170 as an output signal.
- the output circuit 170 outputs an output signal as a signal to the outside of the abnormality detection device 100.
- an actuator driving signal, a data signal for the server, a data signal for another control device or a data processing device, and the like can be considered.
- FIG. 2 is a diagram showing a configuration of software stored in the ROM 120.
- the ROM 120 stores a processing unit 121 and an abnormality detection unit 122 as software executed by the CPU 110.
- these software may be described as an operation subject, but it is added that the CPU 110 actually executes these software.
- the processing unit 121 uses the input signal stored in the RAM 130 to calculate, for example, a control command value and a data processing result.
- the processing unit 121 outputs the calculation result as the output signal described in FIG. 1 via the RAM 130 and the data bus 140.
- the abnormality detection unit 122 detects a processing abnormality of the processing unit 121 using parameters calculated by the processing unit 121.
- the parameter that the abnormality detection unit 122 detects the processing abnormality may be a calculation result (that is, an output signal) by the processing unit 121, or a parameter that is calculated in the middle of obtaining the calculation result.
- a determination result (normal or abnormal) by the abnormality detection unit 122 is output as a signal to the outside through the RAM 130, the input / output port 160, and the output circuit 170.
- FIG. 3 is a functional block diagram showing a detailed configuration of the abnormality detection unit 122.
- the abnormality detection unit 122 includes a data divider 1221, a range setting unit 1222, and an abnormality determination unit 1223. Details of these arithmetic units will be further described.
- FIG. 4 is a diagram for explaining the internal processing of the data divider 1221.
- the data divider 1221 divides the data string obtained by the calculation by the processing unit 121 into one or more clusters.
- a specific clustering method for example, a k-means method can be used. Since the details of the k-means method are generally well known, they are not specifically mentioned here.
- the data divider 1221 outputs cluster data as a result of clustering.
- the cluster data can describe, for example, (a) the cluster number to which each data in the data string belongs, (b) the average value (center vector) of the data belonging to each cluster, and the like.
- FIG. 5 is a diagram illustrating the result of clustering by the data divider 1221.
- the calculation result by the processing unit 121 includes one or more data types. For example, (a) the result of calculating data related to temperature, (b) the result of calculating data related to speed, and the like. By associating these data types with each dimension of the vector, the calculation result by the processing unit 121 can be regarded as vector data.
- the data divider 1221 sets a cluster boundary for each dimension of the vector data.
- the processing by the processing unit 121 as the vector data may be performed by the processing unit 121 itself or the data divider 1221.
- the data divider 1221 acquires a data string (for example, a plurality of calculation results along a time series) for each data type included in the calculation result by the processing unit 121, and for each data type. Define the boundaries of the cluster.
- FIG. 6 is a diagram for explaining the internal processing of the range setter 1222.
- the range setting unit 1222 sets a range in which the calculation result by the processing unit 121 is regarded as normal using the data strings clustered by the data divider 1221. An operation result included in a normal range corresponding to any cluster is regarded as normal.
- the range setter 1222 sets (a) the minimum value in each dimension of the data string belonging to each cluster as the lower limit value in the dimension of the normal range corresponding to the cluster, and (b) each cluster. Is set as the upper limit value in the dimension of the normal range corresponding to the cluster.
- the range setter 1222 outputs range data as a result of setting the normal range.
- the range data describes the lower limit value and the upper limit value of each dimension that defines the normal range corresponding to each cluster.
- FIG. 7 is a diagram showing a result of setting the normal range by the range setter 1222 for the clustering result shown in FIG.
- the upper and lower limits of the normal range are set for each cluster generated by the data divider 1221.
- FIG. 8 is a diagram for explaining the internal processing of the abnormality determiner 1223.
- the abnormality determiner 1223 acquires a new calculation result from the processing unit 121 after the range setting unit 1222 sets the normal range, and whether or not the newly acquired calculation result is normal based on the set normal range. Determine whether.
- the abnormality determiner 1223 performs the abnormality determination by the following process.
- a cluster closest to the newly obtained calculation result is specified. Specifically, among the center vectors (average values of each dimension) of each cluster, the closest one to the operation result newly acquired in the vector space is specified.
- the abnormality detection apparatus 100 sets a normal range by clustering calculation results obtained by the processing unit 121, and determines that the calculation result is normal when a new calculation result is included in the normal range. Thereby, when the software (namely, the process part 121) which CPU110 performs has produced process abnormality, the abnormality can be detected and the reliability of the system provided with the said software can be improved.
- the data string used when the data divider 1221 and the range setter 1222 set the normal range is, for example, a point in time when the processing unit 121 is considered to be operating normally (there is a record of normal operation).
- the result of the calculation in (3) can be used.
- the data to be subjected to abnormality determination by the abnormality determiner 1223 is, for example, a calculation result at a time when there is a possibility that the processing unit 121 is operating abnormally (when there is no record of normal operation). As an example, this corresponds to a case where the processing content executed by the processing unit 121 is updated. The same applies to the following embodiments.
- FIG. 9 is a diagram illustrating a configuration of software stored in the ROM 120 according to the second embodiment.
- the software configuration is the same as that of the first embodiment, the abnormality detection unit 122 receives a data string given by an external signal as a data string used for setting a normal range. Since the other configuration of the abnormality detection device 100 is substantially the same as that of the first embodiment, the following description will focus on the differences.
- the external signal used by the abnormality detection unit 122 is written as data to the RAM 130 via the input circuit 150 and the like as in the first embodiment.
- a detection signal output from a sensor, a data signal from a server, a data signal from another control device or a data processing device, and the like can be considered as an external signal.
- FIG. 10 is a functional block diagram illustrating a detailed configuration of the abnormality detection unit 122 according to the second embodiment.
- the abnormality detection unit 122 has the same configuration as that of the first embodiment, but the data divider 1221 and the range setting unit 1222 are configured to perform clustering and normal range setting using a data string supplied from the outside, respectively. Different from 1.
- the contents of the processing performed by the data divider 1221, the range setter 1222, and the abnormality determiner 1223 are the same as those in the first embodiment. That is, the abnormality determination unit 1223 determines that the calculation result by the processing unit 121 is normal when the calculation result is within the normal range set by the data divider 1221 and the range setting unit 1222, and determines that the calculation result is abnormal when the calculation result is not within the normal range. To do.
- the abnormality detection apparatus 100 can set a normal range in advance using a data string provided from the outside, and can detect an abnormality in the processing unit 121 according to the normal range. Thereby, the same effect as Embodiment 1 can be exhibited.
- a data string used when the normal range is set in advance for example, a data string output from an apparatus of the same type as the abnormality detection apparatus 100 and the processing unit 121 and having a normal operation record can be considered.
- a data string provided from the outside and a data string output from the processing unit 121 can be used in combination as a data string used when setting the normal range.
- these data strings may be simply added to increase the amount of data, or one of the data strings may be selected according to the timing at which processing is performed.
- FIG. 11 is a functional block diagram illustrating a detailed configuration of the abnormality detection unit 122 according to the third embodiment.
- the abnormality detection unit 122 includes a temporal pattern frequency calculator 1224 in addition to the configuration described in the first embodiment. Since the other configuration of the abnormality detection device 100 is substantially the same as that of the first embodiment, the following description will focus on the differences.
- the data divider 1221 and the range setting unit 1222 perform clustering and normal range setting using the data string output from the processing unit 121, as in the first embodiment.
- the data string output by the processing unit 121 is time-series data in which values obtained each time the CPU 110 executes the processing unit 121 are described over time corresponding to the execution time.
- the temporal pattern frequency calculator 1224 identifies a temporal transition pattern in which individual data included in the data string output by the processing unit 121 transitions between clusters with the passage of time, and the occurrence frequency is determined for each pattern. calculate. A specific example will be described later with reference to FIG.
- FIG. 12 is a diagram for explaining the internal processing of the temporal pattern frequency calculator 1224.
- the temporal pattern frequency calculator 1224 calculates the temporal transition pattern of the data string and its occurrence frequency according to the following procedure.
- the temporal pattern frequency calculator 1224 includes a cluster k to which the data k belongs and a cluster to which the data k + 1 belongs. Each k + 1 is specified.
- a procedure for specifying a cluster to which data belongs for example, as described in the first embodiment, it may be possible to specify a cluster having the closest center vector. With this step, a temporal transition pattern in which data transitions from cluster k to cluster k + 1 is specified.
- the temporal pattern frequency calculator 1224 obtains the number of times each data in the data string output from the processing unit 121 has changed from the cluster k to the cluster k + 1.
- the temporal pattern frequency calculator 1224 similarly calculates the temporal transition pattern and its occurrence frequency for other data in the data string output by the processing unit 121.
- FIG. 13 is a diagram illustrating an example of the frequency of occurrence of a temporal transition pattern calculated by the temporal pattern frequency calculator 1224.
- the temporal pattern frequency calculator 1224 may calculate a temporal transition pattern between clusters, or may calculate a temporal transition pattern between normal ranges corresponding to each cluster instead. Here, an example in which a temporal transition pattern between normal ranges is calculated is shown.
- the temporal pattern frequency calculator 1224 may calculate the frequency of occurrence by a method that is easy to calculate, for example, by dividing the frequency of occurrence of the temporal transition pattern by the number of data updates.
- FIG. 14 is a diagram illustrating the internal processing of the abnormality determiner 1223 according to the third embodiment.
- the abnormality determination unit 1223 obtains new time-series data from the processing unit 121 after the temporal pattern frequency calculator 1224 calculates the occurrence frequency described in FIG. 12 in advance, and the new time-series data according to the following procedure. It is determined whether or not is normal.
- FIG. 14 Determination procedure step 1
- the abnormality determiner 1223 uses the same procedure as in step 1 described with reference to FIG. 12 for the data k and data k + 1 (k is a subscript indicating time) in the new data string output by the processing unit 121. And cluster k + 1 are respectively identified.
- the abnormality determiner 1223 obtains the number of times each data in the data string output from the processing unit 121 has changed over time from the cluster k to the cluster k + 1 (or from the normal range k corresponding to each cluster to the normal range k + 1). The abnormality determiner 1223 determines that the temporal transition pattern is normal if the frequency of occurrence of the temporal transition pattern is equal to or higher than a predetermined threshold, and determines that it is abnormal if the frequency of the temporal transition pattern is less than the threshold.
- the threshold value used by the abnormality determiner 1223 can be determined according to the occurrence frequency of each temporal transition pattern calculated in advance by the temporal pattern frequency calculator 1224 according to the procedure described with reference to FIGS. For example, the smallest (or a slightly smaller value) among the occurrence frequencies of all the temporal transition patterns can be adopted as the threshold value.
- the threshold value may be individually set for each temporal transition pattern. In this case, the frequency of occurrence of each temporal transition pattern (or a value slightly smaller than that) can be adopted as each threshold value.
- the determination threshold value in the range 1 ⁇ the range 1 is set smaller than the determination threshold value in the range 1 ⁇ the range 2.
- the abnormality detection apparatus 100 specifies a time-dependent transition pattern of time-series data and its occurrence frequency in advance, and the time-dependent transition pattern of new time-series data and its occurrence frequency do not match this. If it is, it is determined to be abnormal. Thus, even when individual data is within the normal range described in the first embodiment, this can be detected when the change over time is abnormal.
- the accuracy of abnormality determination can be improved by using the abnormality determination method described in the third embodiment together with other embodiments. As a result, the reliability of the software can be further increased.
- FIG. 15 is a diagram showing a software configuration stored in the ROM 120 according to the fourth embodiment of the present invention.
- the abnormality detection unit 122 uses a control command value that is calculated by the processing unit 121 to control the control target (for example, the actuator) 200 as a data string used to set the normal range. Is used. Since the other configuration of the abnormality detection device 100 is substantially the same as that of the first embodiment, the following description will focus on the differences.
- FIG. 16 is a functional block diagram illustrating a detailed configuration of the abnormality detection unit 122 according to the fourth embodiment.
- the abnormality detection unit 122 has the same configuration as that of the first embodiment, but differs from the first embodiment in that clustering / normal range setting / abnormality determination is performed using a control command value output from the processing unit 121.
- the abnormality detection apparatus 100 performs the same abnormality determination as that of the first embodiment using the control command value output from the processing unit 121. Thereby, it can be detected that the newly obtained control command value is not within the normal range. Therefore, an abnormality in the control software itself before the control object 200 receives the control command value can be detected, so that an abnormality that has occurred in the control object and an abnormality that has occurred in the control software can be distinguished.
- FIG. 17 is a functional block diagram illustrating a detailed configuration of the abnormality detection unit 122 according to the fifth embodiment of the present invention.
- the abnormality detection unit 122 has the same configuration as that of the first embodiment, but the abnormality detection unit 122 uses the output signal of the sensor included in the control target 200 or the sensor output signal as a data string used to set the normal range.
- the processing unit 121 performs a predetermined process. Since the other configuration of the abnormality detection device 100 is substantially the same as that of the first embodiment, the following description will focus on the differences.
- the control target 200 may be provided with a sensor related to the control amount in order to provide feedback used when the processing unit 121 calculates a control command value, for example.
- the processing unit 121 can calculate a control command value using an output signal from the sensor.
- the processing unit 121 can further perform an operation for signal processing on the output signal from the sensor.
- the processing performed on the sensor signal includes, for example, filter processing for noise removal, physical quantity conversion, compensation processing (learning processing) accompanying sensor characteristic change, and the like.
- the abnormality detection unit 122 performs abnormality determination by the method described in the first embodiment on the sensor output signal or the data string obtained as a result of performing the operation on the sensor output signal. In any case, these data strings are used internally by the processing unit 121.
- the abnormality detection apparatus 100 performs the same abnormality determination as that of the first embodiment using the sensor output signal used by the processing unit 121 or the result of processing the sensor output signal. As a result, it is possible to detect a sensor signal value used internally by the processing unit 121 or an abnormality in processing. That is, it is possible to detect an abnormality related to a part where the control software processes the sensor signal.
- FIG. 18 is a diagram illustrating a specific example of the control target 200 according to the sixth embodiment of the present invention.
- the controlled object 200 in the sixth embodiment is an internal combustion engine such as an engine.
- the processing unit 121 includes (a) a target fuel injection amount of the fuel injection valve 201, (b) a target air amount of the electronic throttle 202, and (c) a target ignition timing of the spark plug 203 as control command values for the functional units included in the internal combustion engine. , Is calculated. Other configurations are the same as those of the fourth embodiment.
- the abnormality detection apparatus 100 uses the control command values (air amount, fuel injection amount, ignition timing) of the internal combustion engine to perform abnormality determination similar to that in the first embodiment. Therefore, since it can be detected that the newly obtained control command value (air amount, fuel injection amount, ignition timing) is not within the normal range, it is possible to detect an abnormality in the engine control software. Furthermore, the reliability of the engine control system including the engine control software can be improved.
- FIG. 19 is a diagram illustrating a specific example of the control target 200 according to the seventh embodiment of the present invention.
- the controlled object 200 in the seventh embodiment is a rolling process of a steel plant.
- the processing unit 121 calculates (a) target tension, (b) target reduction position, and (c) target rolling material moving speed as control command values for the rolling process.
- Other configurations are the same as those of the fourth embodiment.
- the abnormality detection apparatus 100 performs the same abnormality determination as that of the first embodiment using the control command values (tension, reduction position, rolling material moving speed) of the rolling process. As a result, it is possible to detect that the newly obtained control command values (tension, reduction position, rolling material moving speed) are not within the normal range, and thus it is possible to detect an abnormality in the rolling process control software. Furthermore, the reliability of a rolling process and a steel plant provided with the rolling process control software can be improved.
- FIG. 20 is a functional block diagram illustrating a detailed configuration of the abnormality detection unit 122 according to the eighth embodiment.
- the abnormality detection unit 122 according to the eighth embodiment includes a verifier 1225 in addition to the configuration described in the first embodiment. Since the other configuration is substantially the same as that of the first embodiment, the following description will focus on the differences related to the verifier 1225.
- the verifier 1225 receives a data string (data A) used when the data divider 1221 / range setting unit 1222 performs clustering and normal range setting, and stores the data string in a storage device such as the RAM 130, for example.
- the verifier 1225 acquires a new calculation result (data B) from the processing unit 121 after the data divider 1221 / range setting unit 1222 performs clustering and normal range setting.
- the verifier 1225 notifies that the data B is normal when the data B matches any of the data A, and notifies that it is abnormal when the data B does not match any of the data A.
- the notification result can be stored in the RAM 130, for example, or can be output as an output signal to the outside.
- the operation of the abnormality detection device 100 described in the first to seventh embodiments is verified using data with obvious features (the data A) in advance, for example. Can do. Therefore, the reliability of the software corresponding to the abnormality detection apparatus 100 and the processing unit 121 can be improved.
- FIG. 21 is a diagram illustrating the abnormality detection device 100 and its peripheral configuration according to Embodiment 9 of the present invention.
- the processing unit 121 and the abnormality detection unit 122 are not necessarily configured on the same device, and can be configured as separate devices.
- the processing unit 121 is configured on the terminal device 300, and the abnormality detection device 100 (that is, the abnormality detection unit 122) acquires a calculation result by the processing unit 121 from the terminal device 300.
- the abnormality detection device 100 and the terminal device 300 are connected by an appropriate communication path, and the abnormality detection device 100 acquires the calculation result from the processing unit 121 via the communication path and stores it in, for example, the RAM 130. Then, the abnormality detection unit 122 can acquire the calculation result via the RAM 130.
- the present invention is not limited to the above embodiment, and includes various modifications.
- the above-described embodiment has been described in detail for easy understanding of the present invention, and is not necessarily limited to one having all the configurations described.
- a part of the configuration of an embodiment can be replaced with the configuration of another embodiment, and the configuration of another embodiment can be added to the configuration of an embodiment.
- each arithmetic unit included in the abnormality detection apparatus 100 can be configured in a cloud computing environment.
- the data divider 1221 and the range setting unit 1222 can be built on the cloud, and the processing unit 121 can be built on one or more terminal devices that use the cloud.
- the abnormality determination device 1223 can be constructed on each terminal device, and each terminal device can detect an abnormality of its processing unit 121.
- a plurality of terminal devices 300 are provided, and each terminal device 300 includes a processing unit 121 and an abnormality determination unit 1223, and may use a data divider 1221 and a range setting unit 1222 constructed on the cloud.
- the above components, functions, processing units, processing means, etc. may be realized in hardware by designing some or all of them, for example, with an integrated circuit.
- Each of the above-described configurations, functions, and the like may be realized by software by interpreting and executing a program that realizes each function by the processor.
- Information such as programs, tables, and files for realizing each function can be stored in a recording device such as a memory, a hard disk, an SSD (Solid State Drive), or a recording medium such as an IC card, an SD card, or a DVD.
- Abnormality detection device 110 CPU 120: ROM 121: Processing unit 122: Abnormality detection unit 1221: Data divider 1222: Range setting unit 1223: Abnormality determination unit 1224: Temporal pattern frequency calculator 1225: Verifier 130: RAM 140: data bus 150: input circuit 160: input / output port 170: output circuit 200: control target 300: terminal device
Abstract
Description
図1は、本発明の実施形態1に係る異常検知装置100の構成を説明するブロック図である。ここでは異常検知装置100が実行するソフトウェアの異常を異常検知装置100自身が検知する構成例を説明するが、これに限るものではない。
本実施形態1に係る異常検知装置100は、処理部121による演算結果をクラスタリングすることにより正常範囲を設定し、新たな演算結果がその正常範囲内に含まれる場合は正常であると判定する。これにより、CPU110が実行するソフトウェア(すなわち処理部121)が処理異常を生じさせている場合、その異常を検知して当該ソフトウェアを備えるシステムの信頼性を高めることができる。
実施形態1において、データ分割器1221と範囲設定器1222は、処理部121が出力するデータ列(演算結果)を用いて正常範囲を設定することを説明した。この正常範囲は、処理部121が出力するデータ列に代えて、外部から与えるデータ列によって設定することもできる。本発明の実施形態2では、その具体的な構成例について説明する。
本実施形態2に係る異常検知装置100は、外部から与えられるデータ列を用いてあらかじめ正常範囲を設定しておき、その正常範囲にしたがって処理部121の異常を検出することができる。これにより実施形態1と同様の効果を発揮することができる。あらかじめ正常範囲を設定する際に用いるデータ列としては、例えば異常検知装置100および処理部121と同型の装置であって、正常稼働実績がある装置が出力するデータ列などが考えられる。
実施形態1~2では、処理部121による演算結果がいずれかのクラスタに対応する正常範囲内に収まるか否かによって、その演算結果が異常であるか否かを判定することを説明した。しかし実際の運用環境においては、個々のデータを個別的に見ると正常範囲内に収まっているものの、各データの経時的変化が異常である場合も存在する。そこで本発明の実施形態3では、処理部121が出力するデータ列の経時的変化パターンが正常であるか否かを追加的に判定する構成例を説明する。
経時パターン頻度演算器1224は、処理部121が出力するデータ列内のデータk(kは時刻を表す添え字)がデータk+1へ経時変化したとき、データkが属するクラスタkとデータk+1が属するクラスタk+1をそれぞれ特定する。データが属するクラスタを特定する手順としては、例えば実施形態1で説明したように中心ベクトルが最も近いクラスタを特定することが考えられる。本ステップにより、データがクラスタkからクラスタk+1へ遷移する経時的遷移パターンが特定される。
経時パターン頻度演算器1224は、処理部121が出力するデータ列内の各データがクラスタkからクラスタk+1へ経時変化した回数を求める。経時パターン頻度演算器1224は、処理部121が出力するデータ列内のその他データについても同様に、経時的遷移パターンおよびその発生頻度を演算する。
異常判定器1223は、処理部121が出力する新たなデータ列内のデータkとデータk+1(kは時刻を表す添え字)について、図12で説明したステップ1と同様の手順により、属するクラスタkとクラスタk+1をそれぞれ特定する。
異常判定器1223は、処理部121が出力するデータ列内の各データがクラスタkからクラスタk+1へ(または各クラスタに対応する正常範囲kから正常範囲k+1へ)経時変化した回数を求める。異常判定器1223は、その経時的遷移パターンの発生頻度が所定閾値以上であればその経時的遷移パターンは正常であると判定し、閾値未満であれば異常であると判定する。
本実施形態3に係る異常検知装置100は、時系列データの経時的遷移パターンおよびその発生頻度をあらかじめ特定しておき、新たな時系列データの経時的遷移パターンおよびその発生頻度がこれに合致しない場合は異常であると判定する。これにより、個々のデータが実施形態1で説明した正常範囲内に収まっている場合であっても、その経時的変化が異常である場合に、これを検出することができる。
図15は、本発明の実施形態4におけるROM120が格納しているソフトウェアの構成を示す図である。ソフトウェア構成は実施形態1と同様であるが、異常検知部122は正常範囲を設定するために用いるデータ列として、処理部121が制御対象(例えばアクチュエータ)200を制御するために演算する制御指令値を用いる。異常検知装置100のその他構成は実施形態1と概ね同様であるため、以下では差異点を中心に説明する。
図17は、本発明の実施形態5における異常検知部122の詳細構成を示す機能ブロック図である。異常検知部122は実施形態1と同様の構成を備えるが、異常検知部122は正常範囲を設定するために用いるデータ列として、制御対象200が備えるセンサの出力信号、またはそのセンサ出力信号に対して処理部121が所定の処理を施した結果を用いる。異常検知装置100のその他構成は実施形態1と概ね同様であるため、以下では差異点を中心に説明する。
図18は、本発明の実施形態6における制御対象200の具体例を示す図である。本実施形態6における制御対象200は、エンジンなどの内燃機関である。処理部121は内燃機関が備える機能部に対する制御指令値として、(a)燃料噴射弁201の目標燃料噴射量、(b)電子スロットル202の目標空気量、(c)点火プラグ203の目標点火時期、を演算する。その他構成は実施形態4と同様である。
図19は、本発明の実施形態7における制御対象200の具体例を示す図である。本実施形態7における制御対象200は、鉄鋼プラントの圧延プロセスである。処理部121は圧延プロセスに対する制御指令値として、(a)目標張力、(b)目標圧下位置、(c)目標圧延材移動速度、を演算する。その他構成は実施形態4と同様である。
実施形態1~8においては、処理部121から新たに得られた演算結果が正常範囲内に含まれるか否かに基づき異常判定を実施することを説明した。このことは、新たに得られた演算結果が正常範囲を設定する際に用いたデータ列のうちいずれかと同一であれば、その新たな演算結果は正常とみなすことができることを意味する。そこで本発明の実施形態8では、このようなデータが正常であることを報知する構成例について説明する。
図21は、本発明の実施形態9に係る異常検知装置100およびその周辺構成を示す図である。実施形態1~8において、処理部121と異常検知部122は必ずしも同一の装置上に構成する必要はなく、これらを別装置として構成することもできる。例えば図21に示す構成においては、処理部121は端末装置300上に構成されており、異常検知装置100(すなわち異常検知部122)は端末装置300から処理部121による演算結果を取得する。
本発明は上記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、上記した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、あ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事が可能であり、また、あ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・置換をすることができる。
110:CPU
120:ROM
121:処理部
122:異常検知部
1221:データ分割器
1222:範囲設定器
1223:異常判定器
1224:経時パターン頻度演算器
1225:検証器
130:RAM
140:データバス
150:入力回路
160:入出力ポート
170:出力回路
200:制御対象
300:端末装置
Claims (15)
- プロセッサがソフトウェアを実行することにより得られる演算結果の異常を検知する異常検知装置であって、
複数の前記演算結果を格納する記憶装置から前記複数の演算結果を取得して1以上のグループへ分割する分割器、
各前記グループ内に含まれる前記演算結果が有するデータ値の値範囲を用いて前記演算結果の正常範囲を前記グループごとに設定する範囲設定器、
前記範囲設定器が前記正常範囲を設定した後に新たな前記演算結果を前記記憶装置から取得し、前記正常範囲にしたがって前記新たな演算結果が正常か否かを判定する、異常判定器、
を備えたことを特徴とする異常検知装置。 - 前記分割器は、前記複数の演算結果を用いて1以上の次元を有するベクトルデータを生成し、
前記分割器は、前記ベクトルデータを生成する際に、前記複数の演算結果の種別を前記ベクトルデータの次元に対応付けた上で前記ベクトルデータを生成し、
前記範囲設定器は、前記ベクトルデータの次元ごとに前記正常範囲を設定し、
前記異常判定器は、前記新たな演算結果の種別に対応する前記ベクトルデータ上の次元において前記新たな演算結果が前記正常範囲内に含まれているか否かに基づき、前記新たな演算結果が正常であるか否かを判定する
ことを特徴とする請求項1記載の異常検知装置。 - 前記範囲設定器は、前記分割器が生成した前記ベクトルデータの各前記次元における最小値を、前記正常範囲の対応する次元における下限値として設定し、
前記範囲設定器は、前記分割器が生成した前記ベクトルデータの各前記次元における最大値を、前記正常範囲の対応する次元における上限値として設定する
ことを特徴とする請求項2記載の異常検知装置。 - 前記異常判定器は、前記新たな演算結果が前記ベクトルデータのベクトル空間上で最も近い前記グループを最近接グループとして特定し、
前記異常判定器は、前記最近接グループに対して前記範囲設定器が設定した前記正常範囲内に前記新たな演算結果が含まれる場合は前記新たな演算結果が正常であると判定し、含まれない場合は異常であると判定する
ことを特徴とする請求項2記載の異常検知装置。 - 前記異常判定器は、前記ベクトルデータのベクトル空間上における各前記グループそれぞれの平均値のなかで前記新たな演算結果と最も近いものを前記最近接グループとして特定する
ことを特徴とする請求項4記載の異常検知装置。 - 前記分割器は、前記複数の演算結果として、前記プロセッサが時系列にしたがって前記ソフトウェアを実行することにより得られる時系列データを取得し、
前記異常検知装置は、前記時系列データ内の各時刻における前記演算結果が前記グループ間で遷移する遷移パターンを算出する遷移パターン算出器を備え、
前記異常判定器は、前記新たな演算結果が前記正常範囲内に含まれる場合であっても、前記遷移パターンから逸脱している場合は異常であると判定する
ことを特徴とする請求項1記載の異常検知装置。 - 前記異常検知装置は、前記プロセッサが時系列にしたがって前記ソフトウェアを実行することにより得られる新たな時系列データとして前記新たな演算結果を取得し、
前記遷移パターン算出器は、前記新たな時系列データ内の各時刻における前記新たな演算結果が前記グループ間で遷移する新たな遷移パターンの発生頻度を前記グループ間の遷移経路ごとに算出し、
前記異常判定器は、前記新たな遷移パターンの発生頻度が少なくともいずれかの前記時刻において所定閾値未満である場合は、前記新たな演算結果が前記正常範囲内に含まれる場合であっても異常であると判定する
ことを特徴とする請求項6記載の異常検知装置。 - 前記分割器は、前記複数の演算結果として、前記プロセッサが前記ソフトウェアを実行する途中過程において算出するパラメータを取得する
ことを特徴とする請求項1記載の異常検知装置。 - 前記ソフトウェアは、制御対象を制御するための制御指令値を演算する処理を実装しており、
前記分割器は、前記複数の演算結果として、前記プロセッサが前記ソフトウェアを実行することにより得られる前記制御指令値を取得する
ことを特徴とする請求項1記載の異常検知装置。 - 前記ソフトウェアは、制御対象を制御するための制御指令値を演算する処理を実装しており、
前記分割器は、前記複数の演算結果として、前記制御対象が備えているセンサが出力する信号値または前記信号値に対して加工処理を施した後の加工値を取得する
ことを特徴とする請求項1記載の異常検知装置。 - 前記ソフトウェアは、内燃機関を制御するための制御指令値として、前記内燃機関の空気量、前記内燃機関の燃料噴射量、および前記内燃機関の点火時期を演算する処理を実装しており、
前記分割器は、前記複数の演算結果として、前記プロセッサが前記ソフトウェアを実行することにより算出する前記空気量、前記燃料噴射量、および前記点火時期を取得する
ことを特徴とする請求項1記載の異常検知装置。 - 前記ソフトウェアは、鉄鋼プラントの圧延プロセスを制御するための制御指令値として、前記圧延プロセスにおける張力、前記圧延プロセスにおける圧下位置、および前記圧延プロセスにおける圧延材移動速度を演算する処理を実装しており、
前記分割器は、前記複数の演算結果として、前記プロセッサが前記ソフトウェアを実行することにより算出する前記張力、前記圧下位置、および前記圧延材移動速度を取得する
ことを特徴とする請求項1記載の異常検知装置。 - 前記異常検知装置は、前記異常判定器による判定結果を報知する報知器を備えた
ことを特徴とする請求項1記載の異常検知装置。 - 前記ソフトウェアは、端末機器の動作を制御する処理を実装しており、
前記分割器は、前記複数の演算結果として、前記端末機器が備えるプロセッサが前記ソフトウェアを実行することにより得られるパラメータを、複数の前記端末機器からそれぞれ取得する
ことを特徴とする請求項1記載の異常検知装置。 - 前記異常判定器は、前記新たな演算結果が前記複数の演算結果のうちいずれかと一致する場合は、前記新たな演算結果が正常であると判定し、いずれとも一致しない場合は異常であると判定する
ことを特徴とする請求項1記載の異常検知装置。
Priority Applications (3)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
PCT/JP2015/078752 WO2017061028A1 (ja) | 2015-10-09 | 2015-10-09 | 異常検知装置 |
JP2017544151A JP6678182B2 (ja) | 2015-10-09 | 2015-10-09 | 異常検知装置 |
US15/765,369 US10956251B2 (en) | 2015-10-09 | 2015-10-09 | Abnormality detection device |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
PCT/JP2015/078752 WO2017061028A1 (ja) | 2015-10-09 | 2015-10-09 | 異常検知装置 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2017061028A1 true WO2017061028A1 (ja) | 2017-04-13 |
Family
ID=58488144
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2015/078752 WO2017061028A1 (ja) | 2015-10-09 | 2015-10-09 | 異常検知装置 |
Country Status (3)
Country | Link |
---|---|
US (1) | US10956251B2 (ja) |
JP (1) | JP6678182B2 (ja) |
WO (1) | WO2017061028A1 (ja) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2019003238A (ja) * | 2017-06-12 | 2019-01-10 | 富士電機株式会社 | 異常検出装置及び異常検出システム |
WO2019207767A1 (ja) * | 2018-04-27 | 2019-10-31 | 株式会社日立製作所 | 制御装置および制御方法 |
WO2020234961A1 (ja) * | 2019-05-20 | 2020-11-26 | 三菱電機株式会社 | 状態推定装置および状態推定方法 |
KR20210022720A (ko) * | 2018-06-21 | 2021-03-03 | 아베엘 리스트 게엠베하 | 기술 시스템 캘리브레이션 방법 |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110245132B (zh) * | 2019-06-12 | 2023-10-31 | 腾讯科技(深圳)有限公司 | 数据异常检测方法、装置、计算机可读存储介质和计算机设备 |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2010095314A1 (ja) * | 2009-02-17 | 2010-08-26 | 株式会社日立製作所 | 異常検知方法及び異常検知システム |
JP2013218725A (ja) * | 2013-06-19 | 2013-10-24 | Hitachi Ltd | 異常検知方法及び異常検知システム |
JP2015088154A (ja) * | 2014-02-10 | 2015-05-07 | 株式会社日立パワーソリューションズ | ヘルスマネージメントシステム及びヘルスマネージメント方法 |
Family Cites Families (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP3147586B2 (ja) * | 1993-05-21 | 2001-03-19 | 株式会社日立製作所 | プラントの監視診断方法 |
US5406502A (en) * | 1993-06-29 | 1995-04-11 | Elbit Ltd. | System and method for measuring the operation of a device |
US6216066B1 (en) * | 1998-07-01 | 2001-04-10 | General Electric Company | System and method for generating alerts through multi-variate data assessment |
US7260501B2 (en) * | 2004-04-21 | 2007-08-21 | University Of Connecticut | Intelligent model-based diagnostics for system monitoring, diagnosis and maintenance |
JP5868216B2 (ja) | 2012-02-27 | 2016-02-24 | 三菱電機株式会社 | クラスタリング装置及びクラスタリングプログラム |
US9355007B1 (en) * | 2013-07-15 | 2016-05-31 | Amazon Technologies, Inc. | Identifying abnormal hosts using cluster processing |
JP6200833B2 (ja) | 2014-02-28 | 2017-09-20 | 株式会社日立製作所 | プラントと制御装置の診断装置 |
JP2016181055A (ja) * | 2015-03-23 | 2016-10-13 | 株式会社東芝 | 情報処理装置 |
US10592328B1 (en) * | 2015-03-26 | 2020-03-17 | Amazon Technologies, Inc. | Using cluster processing to identify sets of similarly failing hosts |
US10360093B2 (en) * | 2015-11-18 | 2019-07-23 | Fair Isaac Corporation | Detecting anomalous states of machines |
US10223191B2 (en) * | 2016-07-20 | 2019-03-05 | International Business Machines Corporation | Anomaly detection in performance management |
US10628289B2 (en) * | 2018-03-26 | 2020-04-21 | Ca, Inc. | Multivariate path-based anomaly prediction |
JP7029362B2 (ja) * | 2018-08-16 | 2022-03-03 | 三菱重工業株式会社 | 異常検出装置、異常検出方法、及びプログラム |
-
2015
- 2015-10-09 JP JP2017544151A patent/JP6678182B2/ja active Active
- 2015-10-09 US US15/765,369 patent/US10956251B2/en active Active
- 2015-10-09 WO PCT/JP2015/078752 patent/WO2017061028A1/ja active Application Filing
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2010095314A1 (ja) * | 2009-02-17 | 2010-08-26 | 株式会社日立製作所 | 異常検知方法及び異常検知システム |
JP2013218725A (ja) * | 2013-06-19 | 2013-10-24 | Hitachi Ltd | 異常検知方法及び異常検知システム |
JP2015088154A (ja) * | 2014-02-10 | 2015-05-07 | 株式会社日立パワーソリューションズ | ヘルスマネージメントシステム及びヘルスマネージメント方法 |
Cited By (14)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2019003238A (ja) * | 2017-06-12 | 2019-01-10 | 富士電機株式会社 | 異常検出装置及び異常検出システム |
WO2019207767A1 (ja) * | 2018-04-27 | 2019-10-31 | 株式会社日立製作所 | 制御装置および制御方法 |
US11795888B2 (en) | 2018-06-21 | 2023-10-24 | Avl List Gmbh | Method for calibrating a technical system |
KR102637278B1 (ko) | 2018-06-21 | 2024-02-15 | 아베엘 리스트 게엠베하 | 기술 시스템 캘리브레이션 방법 |
KR20210022720A (ko) * | 2018-06-21 | 2021-03-03 | 아베엘 리스트 게엠베하 | 기술 시스템 캘리브레이션 방법 |
JP7393363B2 (ja) | 2018-06-21 | 2023-12-06 | アー・ファウ・エル・リスト・ゲゼルシャフト・ミト・ベシュレンクテル・ハフツング | 技術システムの校正方法 |
JP2021527776A (ja) * | 2018-06-21 | 2021-10-14 | アー・ファウ・エル・リスト・ゲゼルシャフト・ミト・ベシュレンクテル・ハフツング | 技術システムの校正方法 |
TWI747055B (zh) * | 2019-05-20 | 2021-11-21 | 日商三菱電機股份有限公司 | 狀態推斷裝置以及狀態推斷方法 |
CN113826128A (zh) * | 2019-05-20 | 2021-12-21 | 三菱电机株式会社 | 状态推定装置及状态推定方法 |
KR102409548B1 (ko) | 2019-05-20 | 2022-06-15 | 미쓰비시덴키 가부시키가이샤 | 상태 추정 장치 및 상태 추정 방법 |
CN113826128B (zh) * | 2019-05-20 | 2023-08-08 | 三菱电机株式会社 | 状态推定装置及状态推定方法 |
KR20210141757A (ko) * | 2019-05-20 | 2021-11-23 | 미쓰비시덴키 가부시키가이샤 | 상태 추정 장치 및 상태 추정 방법 |
JPWO2020234961A1 (ja) * | 2019-05-20 | 2021-09-27 | 三菱電機株式会社 | 状態推定装置および状態推定方法 |
WO2020234961A1 (ja) * | 2019-05-20 | 2020-11-26 | 三菱電機株式会社 | 状態推定装置および状態推定方法 |
Also Published As
Publication number | Publication date |
---|---|
JPWO2017061028A1 (ja) | 2018-07-05 |
US20180285183A1 (en) | 2018-10-04 |
JP6678182B2 (ja) | 2020-04-08 |
US10956251B2 (en) | 2021-03-23 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
WO2017061028A1 (ja) | 異常検知装置 | |
CN107636619B (zh) | 信息处理装置、信息处理系统、信息处理方法及记录介质 | |
US20200150159A1 (en) | Anomaly detection device, anomaly detection method, and storage medium | |
US10783248B2 (en) | Determining an aspect of behavior of an embedded device such as, for example, detecting unauthorized modifications of the code and/or behavior of an embedded device | |
US9535419B2 (en) | Device and method for the residual analysis of a residuum to detect system errors in the system behaviour of a system of an aircraft | |
EP3795975B1 (en) | Abnormality sensing apparatus, abnormality sensing method, and abnormality sensing program | |
US11526783B2 (en) | Abnormality determination device, learning device, and abnormality determination method | |
US10930365B2 (en) | Artificial intelligence based monitoring of solid state drives and dual in-line memory modules | |
US9104199B2 (en) | Engine monitoring | |
US20230075229A1 (en) | Model Update Device, Method, and Program | |
JP2013196698A (ja) | システム監視 | |
US10776530B2 (en) | Methods and apparatus to generate an aircraft system model using a standardized architecture | |
JP6845703B2 (ja) | プラント状況情報提示システム及びプラント状況情報提示方法 | |
JP5948998B2 (ja) | 異常診断装置 | |
WO2019207767A1 (ja) | 制御装置および制御方法 | |
JP6347771B2 (ja) | 異常診断装置、異常診断方法及び異常診断プログラム | |
JP6115607B2 (ja) | 異常診断装置、異常診断方法及び異常診断プログラム | |
US11378943B2 (en) | Information processing device, information processing method, and program | |
JP2009003686A (ja) | ソフトウェア動作解析装置、ソフトウェア動作解析方法、及びソフトウェア動作解析用プログラム | |
JP2020161086A (ja) | 制御装置および補正方法 | |
WO2020092327A1 (en) | Root cause analysis for space weather events | |
WO2024047859A1 (ja) | 異常検知装置、異常検知方法、および、異常検知プログラム | |
JP2020061067A5 (ja) | ||
WO2020178938A1 (ja) | 情報処理装置、情報処理方法及び情報処理プログラム | |
JP6678843B1 (ja) | 異常部分検知装置、異常部分検知方法及びプログラム |
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: 15905847 Country of ref document: EP Kind code of ref document: A1 |
|
ENP | Entry into the national phase |
Ref document number: 2017544151 Country of ref document: JP Kind code of ref document: A |
|
WWE | Wipo information: entry into national phase |
Ref document number: 15765369 Country of ref document: US |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 15905847 Country of ref document: EP Kind code of ref document: A1 |