WO2020261875A1 - 予測システム - Google Patents
予測システム Download PDFInfo
- Publication number
- WO2020261875A1 WO2020261875A1 PCT/JP2020/021298 JP2020021298W WO2020261875A1 WO 2020261875 A1 WO2020261875 A1 WO 2020261875A1 JP 2020021298 W JP2020021298 W JP 2020021298W WO 2020261875 A1 WO2020261875 A1 WO 2020261875A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- operating state
- prediction
- attribute information
- target device
- history
- Prior art date
Links
- 238000000605 extraction Methods 0.000 claims abstract description 17
- 239000000284 extract Substances 0.000 claims abstract description 8
- 230000010365 information processing Effects 0.000 claims description 7
- 238000010801 machine learning Methods 0.000 claims description 5
- 230000006870 function Effects 0.000 claims description 2
- 238000012545 processing Methods 0.000 description 30
- 238000004891 communication Methods 0.000 description 13
- 239000000446 fuel Substances 0.000 description 7
- 238000000034 method Methods 0.000 description 7
- 238000005259 measurement Methods 0.000 description 6
- 238000010586 diagram Methods 0.000 description 5
- CIWBSHSKHKDKBQ-JLAZNSOCSA-N Ascorbic acid Chemical compound OC[C@H](O)[C@H]1OC(=O)C(O)=C1O CIWBSHSKHKDKBQ-JLAZNSOCSA-N 0.000 description 3
- 238000001514 detection method Methods 0.000 description 3
- 238000012423 maintenance Methods 0.000 description 3
- 238000004519 manufacturing process Methods 0.000 description 3
- 239000002994 raw material Substances 0.000 description 3
- 230000006399 behavior Effects 0.000 description 2
- 238000005401 electroluminescence Methods 0.000 description 2
- 230000004044 response Effects 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 1
- 238000002485 combustion reaction Methods 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 230000001537 neural effect Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000000737 periodic effect Effects 0.000 description 1
- 230000002093 peripheral effect Effects 0.000 description 1
- 230000000306 recurrent effect Effects 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 239000000126 substance Substances 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
-
- 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
- G05B17/00—Systems involving the use of models or simulators of said systems
- G05B17/02—Systems involving the use of models or simulators of said systems electric
-
- 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
-
- 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
- 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
- G05B23/0243—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 model based detection method, e.g. first-principles knowledge model
- G05B23/0254—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 model based detection method, e.g. first-principles knowledge model based on a quantitative model, e.g. mathematical relationships between inputs and outputs; functions: observer, Kalman filter, residual calculation, Neural Networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
Definitions
- the present invention relates to a prediction system that predicts the operating state of the target device.
- Patent Document 1 a neural circuit model is generated by learning information on changes in characteristics of various parts in a plurality of plants over time, and based on the degree of similarity to each of the patterns of changes in characteristics over time, A characteristic change prediction system for predicting a future characteristic change pattern of the component over time is disclosed. Further, for example, Patent Document 2 discloses a system in which a similar plant is selected from a plurality of plants and a specific performance index is monitored from operation data.
- Patent No. 27589776 Japanese Unexamined Patent Publication No. 2004-290774
- the future behavior of the target device such as a plant may be affected not only by the past operating conditions but also by factors such as the specifications of the target device and the environment in which the target device is placed. It can be said that these factors are information indicating the category to which the target device belongs when the target device is classified according to various criteria, that is, attribute information.
- an object of the present invention is to provide a prediction system capable of predicting the operating state of the target device in the future in consideration of the attribute information of the target device.
- the prediction system includes a storage unit that stores the history of the operating state of each of the plurality of target devices and the attribute information indicating the attributes of each of the plurality of target devices, and the attribute information of the prediction target device.
- the first acquisition unit that acquires the attribute information filter condition in which at least one attribute information is specified, and the operation state filter condition in which at least one operation state included in the operation state history of the prediction target device is specified is acquired.
- the extraction unit that extracts the history of the operating state of the target device that satisfies the attribute information filter condition and the operating state filter condition among the plurality of target devices by referring to the second acquisition unit and the storage unit, and the extracted operating state. It is characterized by including a prediction unit that predicts the operating state of the prediction target device based on the history.
- the history of the operating state of the target device satisfying the attribute information filter condition for specifying the attribute information of the target device is extracted from the history of the operating state of each of the plurality of target devices, and is based on the history of the operating state.
- the operating state of the device to be predicted is predicted. Therefore, it is possible to predict the future of the operating state in consideration of the attribute information of the target device.
- the present invention it is possible to provide a prediction system capable of predicting the operating state of the target device in the future in consideration of the attribute information of the target device.
- FIG. 1 is a schematic view showing an example of the configuration of the prediction system 1 according to the embodiment of the present invention.
- the prediction system 1 has a server 10 and at least one plant 20.
- the plant 20 is an example of the “target device”.
- the type of the plant 20 is not particularly limited, and may include an oil plant, a chemical plant, a pharmaceutical plant, a food plant, a paper manufacturing plant, and the like.
- the elements constituting the plant 20 are not particularly limited, and are connected to storage equipment for fuel and raw materials, equipment for using fuel and raw materials, or equipment for processing and processing fuel and raw materials, and each element. It may include a piping system or the like.
- the target device to which the prediction system 1 can be applied is not limited to various plants, but may be any device such as an industrial machine.
- each plant 20 is connected to each other so that information can be communicated with each other via a communication network such as the Internet.
- a communication network such as the Internet.
- each plant 20 may be referred to as “plant 20A”, “plant 20B”, etc., and when each plant 20 is generically referred to, it may be simply referred to as "plant 20".
- the server 10 is an example of an information processing device that manages a history of operating states (history of operating states) of each plant 20.
- the server 10 is composed of, for example, one information processing device, but the server 10 may be composed of a plurality of information processing devices.
- the information processing device is a device capable of executing various types of information processing, such as a computer provided with a processor and a storage area. Each part shown in FIG. 1 can be realized, for example, by using a storage area or by executing a program stored in the storage area by a processor.
- the server 10 has, for example, a server communication unit 11, an operation unit 12, a display unit 13, a storage unit 14, and a processing unit 15.
- the server communication unit 11 has a communication interface circuit for connecting the server 10 to the communication network.
- the server communication unit 11 supplies data such as an operation state history (operation state history) received from each plant 20 to the processing unit 15.
- the operation unit 12 may be any device as long as the server 10 can be operated, for example, a touch panel, key buttons, or the like.
- the user can input characters, numbers, symbols, etc. using the operation unit 12.
- the operation unit 12 When the operation unit 12 is operated by the user, the operation unit 12 generates a signal corresponding to the operation. Then, the generated signal is supplied to the processing unit 15 as a user's instruction.
- the display unit 13 may be any device as long as it can display an image, an image, or the like, and is, for example, a liquid crystal display or an organic EL (Electro-Luminescence) display.
- the display unit 13 displays an image corresponding to the image data supplied from the processing unit 15, an image corresponding to the image data, and the like.
- the storage unit 14 has, for example, at least one of a semiconductor memory, a magnetic disk device, and an optical disk device.
- the storage unit 14 stores a driver program, an operating system program, an application program, data, and the like used for processing by the processing unit 15.
- the storage unit 14 stores a communication device driver program or the like that controls the server communication unit 11 as a driver program.
- the various programs may be installed in the storage unit 14 from a computer-readable portable recording medium such as a CD-ROM or a DVD-ROM using a known setup program or the like.
- the storage unit 14 stores the attribute information table, the operation state history table, etc., which will be described later, as data. Further, the storage unit 14 stores display data of various screens as data. Further, the storage unit 14 temporarily stores data related to a predetermined process.
- FIG. 2 is a diagram showing an example of an attribute information table.
- the attribute information table is a table for managing attribute information for each plant.
- the attribute information of the plant 20 is, for example, information indicating a category (attribute) to which the plant 20 belongs when the plant 20 is classified according to various criteria.
- the "attribute” may be referred to as an "external state", an “incidental state”, an “external incidental state", or the like.
- the attribute information table for example, "plant ID”, “region”, “climate”, “manufacturing time”, “user”, “fuel type”, “model”, “designer”, “maintenance person”, etc.
- the attribute information of the plant 20 is recorded.
- the "plant ID” is identification information (ID) for identifying the plant 20.
- the "region” is information indicating the region where the plant 20 is installed.
- “Region” is information indicating the climate of the region where the plant 20 is installed.
- the “manufacturing time” is information indicating the time when the plant 20 was manufactured, and may be represented by, for example, a year, a year, a month, a date, or the like.
- the "model” is a type of the plant 20 as a machine, and for example, the model, method, type, and "designer” are information indicating a person (individual, company, etc.) who designed the plant 20.
- the "maintenance person” is information indicating a person (individual, company, etc.) who performs maintenance of the plant 20.
- the attribute information table is not limited to the above-mentioned items, and may include other attribute information.
- FIG. 3 is a diagram showing an example of an operation state history table.
- the operation status history table is a table for managing the operation status history for each plant 20.
- the operation state history is represented by a plurality of rectangular cells C for each plant.
- the horizontal axis shows the elapsed time from the reference time.
- the reference time can be arbitrarily set by the administrator or the like, but for example, at the start of operation of the plant 20 or at the time of start-up (at the time of the first start-up, a rest period for inspection, etc. has passed. (Including the time of restarting), etc.
- each cell C shows the operation state of the plant 20 at the elapsed time (may be one time point or a period having a predetermined width).
- the type of the operating state is distinguished by the pattern in the cell C.
- the length of time possessed by one cell C can be arbitrarily set in units such as seconds, minutes, hours, days, and weeks.
- the operating conditions are not limited to these, but for example, during normal operation, during an accident (fuel shortage, appearance of foreign matter, temperature rise, temperature drop, cooler trip, fuel system trip, blast, blackout, etc.) Is occurring), an alarm is being generated (another alarm such as a balancing shoot level alarm is being generated), an operation is stopped, and any other event is being generated.
- the processing unit 15 includes one or more processors and peripheral circuits thereof.
- the processing unit 15 comprehensively controls the overall operation of the server 10, and is, for example, a CPU (Central Processing Unit).
- the processing unit 15 controls the operation of the server communication unit 11 and the like so that various processes of the server 10 are executed in an appropriate procedure based on the program and the like stored in the storage unit 14.
- the processing unit 15 executes processing based on a program (operating system program, driver program, application program, etc.) stored in the storage unit 14. Further, the processing unit 15 can execute a plurality of programs (application programs and the like) in parallel.
- the processing unit 15 includes a collecting unit 151, a first acquisition unit 152a, a second acquisition unit 152b, an extraction unit 153, a prediction unit 154, a display processing unit 155, and the like.
- Each of these units included in the processing unit 15 is a functional module implemented by a program executed on the processor included in the processing unit 15.
- each of these units included in the processing unit 15 may be mounted on the server 2 as an independent integrated circuit, microprocessor, or firmware.
- the collecting unit 151 collects (receives) the operating state history of the plant 20 from each plant 20 and records it in the operating state history table stored in the storage unit 14 or the like.
- the timing at which the collection / recording process of the collection unit 151 is executed is not particularly limited, but for example, a predetermined periodic or aperiodic timing, or an administrator or the like can perform the process via the operation unit 12. It may be when the execution command is input.
- the reception unit 152 receives, for example, various filter conditions input by the user via the operation unit 12.
- the filter condition is a condition that the operation state history of the plant 20 to the plant 20 should satisfy in order to extract (filter) a desired operation state history from the operation state history table.
- the first acquisition unit 152a acquires the attribute information filter condition, which is a condition related to the attribute information of the plant 20.
- the attribute information filter condition is a condition that the plant 20 should satisfy in order to extract (filter) a desired operation state history from the operation state history table, and is at least included in the attribute information of the prediction target device.
- One attribute information is the specified condition.
- the attribute information included in the attribute information filter condition may be, for example, the attribute information listed in the above description of FIG. 2, or any other attribute information.
- the first acquisition unit 152a acquires the attribute information filter condition by receiving, for example, the attribute information filter condition input by the user via the operation unit 12.
- the second acquisition unit 152b acquires the operation state filter condition, which is a condition related to the operation state of the plant 20.
- the operation state filter condition is a condition that the operation state history of the plant 20 should satisfy in order to extract (filter) a desired operation state history from the operation state history table.
- the operation state filter condition at least one operation state included in the operation state history of the prediction target device is specified. Further, in the operation state filter condition, the order of occurrence of each operation state may be specified.
- the extraction unit 153 extracts the operation state history of the plant 20 that satisfies the attribute information filter condition received by the first acquisition unit 152a and the operation state filter condition received by the second acquisition unit 152b with reference to the storage unit 14.
- the duration of the operation state designated as the operation state filter condition (when the operation state is intermittent, each It may be extracted only when the time (which may be the total time of the history of the operating state) is equal to or more than a predetermined threshold value.
- the prediction unit 154 predicts the operation state of the prediction target device (plant) based on the operation state history extracted by the extraction unit 153.
- the method of predicting the operating state by the prediction unit 154 is not particularly limited, and may be, for example, a prediction by statistical analysis, a prediction by a probability density function, a prediction based on Bayesian theory, or the like. More specifically, the prediction unit 154 statistically analyzes the operation state history extracted by the extraction unit 153 (for example, the extracted operation state history is aggregated and normalized for each hour). The probability of occurrence of the operating state for each hour may be calculated. The occurrence probability makes it possible to predict the operating state of the prediction target device (plant) at a certain point in the future.
- the prediction unit 154 executes machine learning using, for example, the operation state history extracted by the extraction unit 153 as learning data to generate a learning model, and learns the history of the operation state of the prediction target device (plant).
- the future operating state of the predicted device may be output. More specifically, for example, the prediction unit 154 inputs the time series data of the operation state history by learning the recurrent neural network (RNN) using the operation state history extracted by the extraction unit 153 as learning data. Then, a learning model that outputs the future operating state may be generated.
- the future operating state makes it possible to predict the operating state of the prediction target device (plant) at a certain point in the future.
- the display processing unit 155 causes the display unit 13 to display various screens based on the display data of the various screens stored in the storage unit 14.
- the plant 20 includes an operation unit 21, various sensors 22, a measurement control system 23, and a plant communication unit 24.
- the operation unit 21 has main devices constituting the plant 20, and includes, for example, various modules such as a combustion chamber and a heat exchange chamber, a piping system connecting each module, and the like.
- the sensors 22 are installed in various places in the driving unit 21, detect various physical quantities of the driving unit 21, and supply the detection results to the measurement control system 23.
- the measurement control system 23 generates an operation state history based on the detection result supplied from the sensor 22. Specifically, the measurement control system 23 determines the operating state of the operating unit 21 in which the sensor 22 is installed by analyzing the detection result supplied from the sensor 22, and then determines the operating state of the determined operating state. Generates an operating state history that is a time-series change. Then, the measurement control system 23 transmits the operation status history of the operation unit 21 to the server 10 via the plant communication unit 24.
- FIG. 4 is a diagram showing an example of an operation flow by the prediction system 1 according to the embodiment.
- FIG. 5 is a diagram showing an example of the screen 500 displayed on the display unit 13.
- FIG. 6 is a diagram showing an example of the screen 600 displayed on the display unit 13.
- the individual plants 20 may be referred to as "plant A", "plant B", and the like.
- the collection unit 151 of the server 10 collects the operation state history of the plant 20 from each plant 20 in advance, and records the operation state history in the operation state history table stored in the storage unit 14. ..
- the first acquisition unit 152a of the server 10 acquires the attribute information filter condition. Specifically, for example, the first acquisition unit 152a acquires the attribute information filter condition by accepting the input of the attribute information filter condition in response to the operation of the operation unit 12 by the user.
- the display processing unit 155 of the server 10 causes the display unit 13 to display the screen 500 shown in FIG. 5, for example, based on the display data stored in the storage unit 14.
- the screen 500 includes a display unit 501 for the attribute information filter condition, a display unit 502 for the operation state filter condition, and a display unit 503 for the operation state history of the extracted plant 20.
- the content of the attribute information filter condition acquired by the first acquisition unit 152a is displayed on the display unit 501.
- the display unit 501 displays the attribute information filter conditions in which the area is "cold region", the fuel is "high moisture”, and the model is "small".
- the extraction unit 153 refers to the attribute information table stored in the storage unit 14, identifies the plant 20 that satisfies the attribute information filter condition acquired in S100, and then identifies the operation state of the specified plant 20. Extract the history from the operation status history table.
- the second acquisition unit 152b of the server 10 receives the designation of the plant 20 as the prediction target device in response to the operation of the operation unit 12 by the user. Specifically, the second acquisition unit 152b accepts input of information (for example, the name of the plant 20, identification information, etc.) for identifying the plant 20 designated by the user as the prediction target device.
- information for example, the name of the plant 20, identification information, etc.
- the second acquisition unit 152b of the server 10 refers to the operation state history table stored in the storage unit 14 and generates an operation state filter condition based on the operation state history of the designated plant 20. , Acquire the operation state filter condition.
- the second acquisition unit 152b may select at least one operating state included in the history of the operating state of the plant 20 as the prediction target device, and use the selected operating state as the operating state filter condition.
- the second acquisition unit 152b may use all the operating states included in the operating state history of the plant 20 as the prediction target device as the operating state filter conditions.
- the display unit 502 of the screen 500 displays the contents of the operation state filter condition generated by the second acquisition unit 152b.
- the display unit 502 displays the content of the operating state filter condition acquired by the second acquisition unit 152b.
- "operating state ⁇ ", "operating state ⁇ ", and "operating state ⁇ " are displayed on the display unit 502 as operating state filter conditions. This is because the operating state history of the plant X includes the history of "operating state ⁇ ", "operating state ⁇ ", and "operating state ⁇ ", respectively.
- the extraction unit 153 extracts the operation state history that satisfies the operation state filter condition generated in S103 from the operation state history extracted in S101.
- the display unit 503 of the screen 500 shows the operation status history of the plant 20 (specifically, plant A, plant C, plant) extracted as satisfying the attribute information filter condition and the operation status filter condition. E and the operation status history of the plant F) are displayed.
- the prediction unit 154 predicts the operation state of the prediction target device based on the operation state history of the plant 20 extracted in S101. Specifically, for example, the prediction unit 154 may calculate the probability of occurrence of the driving state for each hour by statistically analyzing the extracted driving state history. Alternatively, for example, the prediction unit 154 executes machine learning using the operation state history extracted by the extraction unit 153 as learning data to generate a learning model, and learns the history of the operation state of the prediction target device (plant). By inputting to the model, an output indicating the future operating state of the predicted device (plant) may be generated.
- the display processing unit 155 causes the display unit 13 to display the prediction result by the prediction unit 154, and the operation processing of the prediction system 1 is completed.
- FIG. 6A is an example of a screen 600 showing a prediction result displayed on the display unit 13 by the display processing unit 155 when the prediction unit 154 calculates the occurrence probability of the operating state.
- the screen 600 includes a display unit 601 of the occurrence probability of the operation state calculated by the prediction unit 154 and a display unit 602 of the operation state history of the prediction target device.
- the display unit 601 displays the time-series change of the occurrence probability of each operating state calculated by the prediction unit 154.
- FIG. 6A is an example of a screen 600 showing a prediction result displayed on the display unit 13 by the display processing unit 155 when the prediction unit 154 calculates the occurrence probability of the operating state.
- the screen 600 includes a display unit 601 of the occurrence probability of the operation state calculated by the prediction unit 154 and a display unit 602 of the operation state history of the prediction target device.
- the display unit 601 displays the time-series change of the occurrence probability of each operating state calculated by the prediction unit 154.
- the operation state history of the plant X which is the device for predicting the operation state, is displayed. Specifically, the display unit 602 displays the operation status history of the plant X from the start of operation to the elapse of time T1.
- the display unit 602 displays the operation status history of the plant X from the start of operation to the elapse of time T1.
- the probability that the plant X will be in the operating state ⁇ at the future time T2 (> T1) is P ⁇ (T2)
- the probability that it will be in the operating state ⁇ is P ⁇ (T2).
- FIG. 6B is an example of a screen 700 showing a prediction result displayed on the display unit 13 by the display processing unit 155 when the prediction unit 154 makes a prediction by a learning model by machine learning.
- the screen 700 includes a display unit 701 of a time-series change in the operating state of the plant X, which is a prediction target device.
- the display unit 701 displays the history of the operating state of the plant X up to the current time T1.
- the history of the operating state of the plant X up to the current time T1 becomes learning data in machine learning executed by the prediction unit 154.
- the display unit 701 displays the future operation state as an output obtained by inputting the operation state history of the plant X into the learning model generated by the prediction unit 154 using the learning data.
- the second acquisition unit 152b receives the designation of the plant 20 to be the prediction target device (S102), and then sets the operation state filter condition based on the operation state history of the designated plant 20. By generating (S103), it was decided to acquire the operating state filter condition.
- the second acquisition unit 152b may acquire the operation state filter condition input by the user, for example, by operating the operation unit 12.
- Prediction system 10 ... server, 11 ... server communication unit, 12 ... operation unit, 13 ... display unit, 14 ... storage unit, 15 ... processing unit, 151 ... collection unit, 152a ... first acquisition unit, 152b ... 2 Acquisition unit, 153 ... Extraction unit, 154 ... Prediction unit, 155 ... Display processing unit, 20, 20A, 20B, 20C ... Plant, 21 ... Operation unit, 22 ... Sensor, 23 ... Measurement control system, 24 ... Plant communication unit
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Mathematical Physics (AREA)
- Testing And Monitoring For Control Systems (AREA)
Abstract
Description
(1-1)予測システム1
図1は、本発明の実施形態に係る予測システム1の構成の一例を示す概略図である。図1に示すとおり、予測システム1は、サーバ10と、少なくとも1のプラント20とを有する。ここで、プラント20は「対象装置」の一例である。プラント20の種類は特に限定されず、石油プラント、化学プラント、医薬品プラント、食品プラント、及び製紙プラント等を含んでもよい。また、プラント20を構成する要素は、特に限定されず、燃料や原材料等の貯蔵設備、燃料や原材料等を使用、または燃料や原材料等に対して処理・加工等を行う設備、各要素を接続する配管系統等を含んでもよい。なお、予測システム1が適用可能な対象装置は、各種のプラントに限らず、産業機械等の任意の装置であってよい。
サーバ10は、各プラント20の運転状態の履歴(運転状態履歴)を管理する情報処理装置の一例である。本例では、サーバ10は、例えば、1つの情報処理装置によって構成されるものとするが、サーバ10は、複数の情報処理装置によって構成されてもよい。ここで、情報処理装置は、例えば、プロセッサ及び記憶領域を備えたコンピュータ等の、各種の情報処理を実行可能な装置である。図1に示す各部は、例えば、記憶領域を用いたり、記憶領域に格納されたプログラムをプロセッサが実行したりすることにより実現することができる。
プラント20は、運転部21と、各種のセンサ22と、計測制御システム23と、プラント通信部24とを有する。運転部21は、プラント20を構成する主要な装置を有しており、例えば、燃焼室、及び熱交換室等の種々のモジュールや、各モジュールを接続する配管系統等を含む。センサ22は、運転部21の各所に設置され、運転部21の各種の物理量を検知し、検知結果を計測制御システム23に供給する。計測制御システム23は、センサ22から供給される検知結果に基づいて運転状態履歴を生成する。具体的には、計測制御システム23は、センサ22から供給された検知結果を解析することにより、当該センサ22が設置された運転部21の運転状態を判定した上で、判定された運転状態の時系列的な変化である運転状態履歴を生成する。そして、計測制御システム23は、プラント通信部24を介して、運転部21の運転状態履歴を、サーバ10に送信する。
次に、図4~6を参照して、実施形態に係る予測システム1の動作処理の一例について説明する。図4は、実施形態に係る予測システム1による動作フローの一例を示す図である。図5は、表示部13に表示される画面500の一例を示す図である。図6は、表示部13に表示される画面600の一例を示す図である。以下では、個別のプラント20を「プラントA」、「プラントB」などと称する場合がある。
まず、サーバ10の第1取得部152aは、属性情報フィルタ条件を取得する。具体的には、例えば、第1取得部152aは、ユーザによる操作部12の操作に応じて、属性情報フィルタ条件の入力を受け付けることにより、当該属性情報フィルタ条件を取得する。このとき、サーバ10の表示処理部155は、例えば、記憶部14に記憶された表示データに基づいて、図5に示す画面500を表示部13に表示させる。図5に示すとおり、画面500は、属性情報フィルタ条件の表示部501と、運転状態フィルタ条件の表示部502と、抽出されたプラント20の運転状態履歴の表示部503とを含む。表示部501には、第1取得部152aが取得した属性情報フィルタ条件の内容が表示される。図5に示す例では、当該表示部501に、地域が「寒冷地」で、燃料が「水分大」で、機種が「小型」である属性情報フィルタ条件が表示されている。
次に、抽出部153は、記憶部14に記憶された属性情報テーブルを参照して、S100で取得された属性情報フィルタ条件を満たすプラント20を特定した上で、特定されたプラント20の運転状態履歴を運転状態履歴テーブルから抽出する。
次に、サーバ10の第2取得部152bは、ユーザによる操作部12の操作に応じて、予測対象装置となるプラント20の指定を受け付ける。具体的には、第2取得部152bは、予測対象装置としてユーザが指定するプラント20を特定するための情報(例えば、当該プラント20の名称や、識別情報等)の入力を受け付ける。
次に、サーバ10の第2取得部152bは、記憶部14に記憶された運転状態履歴テーブルを参照して、指定されたプラント20の運転状態履歴に基づいて運転状態フィルタ条件を生成することにより、当該運転状態フィルタ条件を取得する。例えば、第2取得部152bは、予測対象装置となるプラント20の運転状態の履歴に含まれる少なくとも1つの運転状態を選択して、選択された運転状態を運転状態フィルタ条件としてもよい。特に、第2取得部152bは、予測対象装置となるプラント20の運転状態の履歴に含まれる全ての運転状態を、運転状態フィルタ条件としてもよい。
次に、抽出部153は、S101で抽出された運転状態履歴から、S103で生成された運転状態フィルタ条件を満たす運転状態履歴を抽出する。
次に、予測部154は、S101で抽出されたプラント20の運転状態履歴に基づいて、予測対象装置の運転状態を予測する。具体的には、例えば、予測部154は、抽出された運転状態履歴を統計分析することにより時間毎の運転状態の発生確率を算出してもよい。或いは、例えば、予測部154は、抽出部153により抽出された運転状態履歴を学習データとする機械学習を実行して学習モデルを生成し、予測対象装置(プラント)の運転状態の履歴を当該学習モデルに入力することにより、予測対象装置(プラント)の将来の運転状態を示す出力を生成してもよい。
次に、表示処理部155は、予測部154による予測結果を表示部13に表示させ、予測システム1の動作処理が終了する。
Claims (14)
- 複数の対象装置それぞれの運転状態の履歴と、前記複数の対象装置それぞれの属性を示す属性情報とを記憶する記憶部と、
予測対象装置の属性情報に含まれる少なくとも1つの属性情報が指定された属性情報フィルタ条件を取得する第1取得部と、
前記予測対象装置の運転状態の履歴に含まれる少なくとも1つの運転状態が指定された運転状態フィルタ条件を取得する第2取得部と、
前記記憶部を参照して、前記属性情報フィルタ条件及び前記運転状態フィルタ条件を満たす対象装置の運転状態の履歴を抽出する抽出部と、
抽出された前記運転状態の履歴に基づいて前記予測対象装置の運転状態を予測する予測部と、
を備えることを特徴とする予測システム。 - 前記第1取得部は、前記属性情報フィルタ条件の入力を受け付けることにより、前記属性情報フィルタ条件を取得する、請求項1に記載の予測システム。
- 前記第2取得部は、前記予測対象装置の指定を受け付け、前記記憶部に記憶された前記予測対象装置の運転状態の履歴に基づいて前記運転状態フィルタ条件を生成する、請求項1又は2に記載の予測システム。
- 前記第2取得部は、前記運転状態フィルタ条件の入力を受け付けることにより、前記属性情報フィルタ条件を取得する、請求項1又は2に記載の予測システム。
- 前記予測部は、前記抽出された前記運転状態の履歴を統計分析して運転状態の発生確率を算出する、請求項1から4のいずれか一項に記載の予測システム。
- 前記予測部が算出した前記発生確率を表示する表示部、を更に備える、請求項5に記載の予測システム。
- 前記予測部は、前記抽出された前記運転状態の履歴を学習データとする機械学習を実行して学習モデルを生成し、前記予測対象装置の運転状態を当該学習モデルに入力することにより、前記予測対象装置の将来の運転状態を出力させる、請求項1から4のいずれか一項に記載の予測システム。
- 前記出力された前記予測対象装置の将来の運転状態を表示する表示部、を更に備える、請求項7に記載の予測システム。
- 前記運転状態フィルタ条件において、複数の運転状態が指定されており、
前記運転状態フィルタ条件は、前記複数の運転状態の発生の順序を含む、請求項1から8のいずれか一項に記載の予測システム。 - 前記対象装置は、プラントである、請求項1から9のいずれ一項に記載の予測システム。
- 前記対象装置は、ボイラである、請求項10に記載の予測システム。
- 対象装置の運転状態を把握するための表示装置であって、
複数の対象装置それぞれの運転状態の履歴及び属性情報から、予測対象装置の属性情報に含まれる少なくとも1つの属性情報が指定された属性情報フィルタ条件と、前記予測対象装置の運転状態の履歴に含まれる少なくとも1つの運転状態が指定された運転状態フィルタ条件とを満たすものとして抽出された対象装置の運転状態の履歴に基づいた前記予測対象装置の運転状態の予測結果を表示する、表示装置。 - 複数の対象装置それぞれの運転状態の履歴と、前記複数の対象装置それぞれの属性を示す属性情報とを記憶する記憶部を備える情報処理装置に、
予測対象装置の属性情報に含まれる少なくとも1つの属性情報が指定された属性情報フィルタ条件を取得するステップと、
前記予測対象装置の運転状態の履歴に含まれる少なくとも1つの運転状態が指定された運転状態フィルタ条件を取得するステップと、
前記記憶部を参照して、前記複数の対象装置のうち前記属性情報フィルタ条件及び前記運転状態フィルタ条件を満たす対象装置の運転状態の履歴を抽出するステップと、
抽出された前記運転状態の履歴に基づいて前記予測対象装置の運転状態を予測するステップと、
を実行させるための方法。 - 複数の対象装置それぞれの運転状態の履歴と、前記複数の対象装置それぞれの属性を示す属性情報とを記憶する記憶部を備える情報処理装置を、
予測対象装置の属性情報に含まれる少なくとも1つの属性情報が指定された属性情報フィルタ条件を取得する第1取得部と、
前記予測対象装置の運転状態の履歴に含まれる少なくとも1つの運転状態が指定された運転状態フィルタ条件を取得する第2取得部と、
前記記憶部を参照して、前記複数の対象装置のうち前記属性情報フィルタ条件及び前記運転状態フィルタ条件を満たす対象装置の運転状態の履歴を抽出する抽出部と、
抽出された前記運転状態の履歴に基づいて前記予測対象装置の運転状態を予測する予測部と、
として機能させるためのプログラム。
Priority Applications (7)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202080047121.5A CN114041095A (zh) | 2019-06-28 | 2020-05-29 | 预测系统 |
BR112021026482A BR112021026482A2 (pt) | 2019-06-28 | 2020-05-29 | Sistema de predição, dispositivo de exibição para identificar um estado de operação de um dispositivo alvo, método de execução de um dispositivo e programa |
JP2021527524A JP7480141B2 (ja) | 2019-06-28 | 2020-05-29 | 予測システム |
AU2020302438A AU2020302438A1 (en) | 2019-06-28 | 2020-05-29 | Prediction system |
KR1020217040477A KR20220027841A (ko) | 2019-06-28 | 2020-05-29 | 예측시스템 |
EP20833158.7A EP3992740A4 (en) | 2019-06-28 | 2020-05-29 | PREDICTION SYSTEM |
US17/561,292 US20220121189A1 (en) | 2019-06-28 | 2021-12-23 | Prediction system |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2019-122012 | 2019-06-28 | ||
JP2019122012 | 2019-06-28 |
Related Child Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US17/561,292 Continuation US20220121189A1 (en) | 2019-06-28 | 2021-12-23 | Prediction system |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2020261875A1 true WO2020261875A1 (ja) | 2020-12-30 |
Family
ID=74061911
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2020/021298 WO2020261875A1 (ja) | 2019-06-28 | 2020-05-29 | 予測システム |
Country Status (9)
Country | Link |
---|---|
US (1) | US20220121189A1 (ja) |
EP (1) | EP3992740A4 (ja) |
JP (1) | JP7480141B2 (ja) |
KR (1) | KR20220027841A (ja) |
CN (1) | CN114041095A (ja) |
AU (1) | AU2020302438A1 (ja) |
BR (1) | BR112021026482A2 (ja) |
TW (1) | TWI744909B (ja) |
WO (1) | WO2020261875A1 (ja) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2023050478A (ja) * | 2021-09-30 | 2023-04-11 | 横河電機株式会社 | 監視装置、監視方法、および監視プログラム |
EP4184267A1 (en) * | 2021-11-22 | 2023-05-24 | Abb Schweiz Ag | Method for determining the state of health of an industrial process |
JP7325695B1 (ja) * | 2023-01-23 | 2023-08-14 | 三菱電機株式会社 | データ処理装置、データ処理方法及びデータ処理プログラム |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2758976B2 (ja) | 1990-06-07 | 1998-05-28 | 株式会社日立製作所 | 特性変化予測方法およびシステム |
JP2001101235A (ja) * | 1999-09-28 | 2001-04-13 | Tokyo Gas Co Ltd | 設備劣化診断方法及び設備劣化診断システム |
JP2004290774A (ja) | 2003-03-26 | 2004-10-21 | Ngk Insulators Ltd | 遠隔監視システム |
JP2006072408A (ja) * | 2004-08-31 | 2006-03-16 | Yokogawa Electric Corp | 設備制御方法、設備制御装置および設備制御のためのプログラム |
WO2018146768A1 (ja) * | 2017-02-09 | 2018-08-16 | 三菱電機株式会社 | 不良要因推定装置および不良要因推定方法 |
Family Cites Families (15)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP4017272B2 (ja) * | 1998-12-16 | 2007-12-05 | 中国電力株式会社 | プラント状態推定・予測装置及び方法 |
CN100429953C (zh) * | 2002-10-10 | 2008-10-29 | 松下电器产业株式会社 | 信息取得方法、信息提供方法及信息取得装置 |
US8457908B2 (en) * | 2009-06-11 | 2013-06-04 | University Of Washington | Sensing events affecting liquid flow in a liquid distribution system |
JP5364530B2 (ja) * | 2009-10-09 | 2013-12-11 | 株式会社日立製作所 | 設備状態監視方法、監視システム及び監視プログラム |
JP5960029B2 (ja) * | 2012-10-31 | 2016-08-02 | 住友重機械工業株式会社 | 異常原因特定システム |
US9405286B2 (en) * | 2013-03-01 | 2016-08-02 | Fisher-Rosemount Systems, Inc. | Use of predictors in process control systems with wireless or intermittent process measurements |
CN103400310B (zh) * | 2013-08-08 | 2017-04-26 | 华北电力大学(保定) | 考虑历史数据趋势预测的配电网电气设备状态评估方法 |
CN104793072A (zh) * | 2015-03-28 | 2015-07-22 | 合肥天海电气技术有限公司 | 基于热图谱分析法的电气设备故障诊断系统 |
US10545120B2 (en) * | 2016-02-23 | 2020-01-28 | John Crane Uk Ltd. | Systems and methods for predictive diagnostics for mechanical systems |
US10211634B2 (en) * | 2016-02-24 | 2019-02-19 | Mitsubishi Electric Research Laboratories, Inc. | Dynamic state estimation of power distribution system |
JP6723669B2 (ja) * | 2016-09-27 | 2020-07-15 | 東京エレクトロン株式会社 | 異常検知プログラム、異常検知方法および異常検知装置 |
RU2649542C1 (ru) * | 2016-12-06 | 2018-04-03 | Акционерное общество "РОТЕК" (АО "РОТЕК") | Способ и система удаленного мониторинга объектов |
CN108491965B (zh) * | 2018-03-14 | 2021-11-05 | 广东省智能制造研究所 | 冲压设备的状态预测方法、装置、电子设备与存储介质 |
CN108680244B (zh) * | 2018-04-26 | 2020-06-09 | 浙江大学 | 一种旋转机械振动无线监测装置及方法 |
CN108763377B (zh) * | 2018-05-18 | 2021-08-13 | 郑州轻工业学院 | 基于卫星故障诊断多源遥测大数据特征提取预处理方法 |
-
2020
- 2020-05-22 TW TW109117078A patent/TWI744909B/zh active
- 2020-05-29 BR BR112021026482A patent/BR112021026482A2/pt unknown
- 2020-05-29 JP JP2021527524A patent/JP7480141B2/ja active Active
- 2020-05-29 CN CN202080047121.5A patent/CN114041095A/zh active Pending
- 2020-05-29 AU AU2020302438A patent/AU2020302438A1/en active Pending
- 2020-05-29 EP EP20833158.7A patent/EP3992740A4/en active Pending
- 2020-05-29 KR KR1020217040477A patent/KR20220027841A/ko unknown
- 2020-05-29 WO PCT/JP2020/021298 patent/WO2020261875A1/ja active Application Filing
-
2021
- 2021-12-23 US US17/561,292 patent/US20220121189A1/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2758976B2 (ja) | 1990-06-07 | 1998-05-28 | 株式会社日立製作所 | 特性変化予測方法およびシステム |
JP2001101235A (ja) * | 1999-09-28 | 2001-04-13 | Tokyo Gas Co Ltd | 設備劣化診断方法及び設備劣化診断システム |
JP2004290774A (ja) | 2003-03-26 | 2004-10-21 | Ngk Insulators Ltd | 遠隔監視システム |
JP2006072408A (ja) * | 2004-08-31 | 2006-03-16 | Yokogawa Electric Corp | 設備制御方法、設備制御装置および設備制御のためのプログラム |
WO2018146768A1 (ja) * | 2017-02-09 | 2018-08-16 | 三菱電機株式会社 | 不良要因推定装置および不良要因推定方法 |
Non-Patent Citations (1)
Title |
---|
See also references of EP3992740A4 |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2023050478A (ja) * | 2021-09-30 | 2023-04-11 | 横河電機株式会社 | 監視装置、監視方法、および監視プログラム |
JP7452509B2 (ja) | 2021-09-30 | 2024-03-19 | 横河電機株式会社 | 監視装置、監視方法、および監視プログラム |
EP4184267A1 (en) * | 2021-11-22 | 2023-05-24 | Abb Schweiz Ag | Method for determining the state of health of an industrial process |
WO2023089085A1 (en) * | 2021-11-22 | 2023-05-25 | Abb Schweiz Ag | Methods for determining the state of health of an industrial process |
JP7325695B1 (ja) * | 2023-01-23 | 2023-08-14 | 三菱電機株式会社 | データ処理装置、データ処理方法及びデータ処理プログラム |
Also Published As
Publication number | Publication date |
---|---|
CN114041095A (zh) | 2022-02-11 |
EP3992740A1 (en) | 2022-05-04 |
AU2020302438A1 (en) | 2022-01-27 |
JP7480141B2 (ja) | 2024-05-09 |
TW202117476A (zh) | 2021-05-01 |
JPWO2020261875A1 (ja) | 2020-12-30 |
EP3992740A4 (en) | 2022-08-17 |
TWI744909B (zh) | 2021-11-01 |
BR112021026482A2 (pt) | 2022-02-08 |
KR20220027841A (ko) | 2022-03-08 |
US20220121189A1 (en) | 2022-04-21 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
WO2020261875A1 (ja) | 予測システム | |
Pandey et al. | A methodology for joint optimization for maintenance planning, process quality and production scheduling | |
CA3007973C (en) | System and method for monitoring manufacturing | |
US7103610B2 (en) | Method, system and computer product for integrating case based reasoning data and failure modes, effects and corrective action data | |
JP6411769B2 (ja) | 状態監視装置 | |
US10699248B2 (en) | Inspection management system and inspection management method | |
JP2006277370A (ja) | 回路基板の品質解析システム及び品質解析方法 | |
JP2005182465A (ja) | 保守支援方法及びプログラム | |
Yang et al. | Machine learning-based prognostics for central heating and cooling plant equipment health monitoring | |
CN103425093B (zh) | 生产工厂中故障状态自动恢复的方法和系统 | |
Liu et al. | Objective-oriented optimal sensor allocation strategy for process monitoring and diagnosis by multivariate analysis in a Bayesian network | |
JP5387779B2 (ja) | 運用管理装置、運用管理方法、及びプログラム | |
JP2018180759A (ja) | システム分析装置、及びシステム分析方法 | |
Pandey et al. | A methodology for simultaneous optimisation of design parameters for the preventive maintenance and quality policy incorporating Taguchi loss function | |
WO2013041440A1 (en) | System and method for plant wide asset management | |
CN108701269A (zh) | 生产设施经营优化装置 | |
EP3598258B1 (en) | Risk assessment device, risk assessment system, risk assessment method, and risk assessment program | |
CN100435052C (zh) | 工序管理装置及其控制方法、工序管理程序及记录介质 | |
Fong et al. | Methods for Assessing Product Reliability: Looking for enhancements by adopting condition-based monitoring | |
EP3614221B1 (en) | Risk evaluation device, risk evaluation system, risk evaluation method, risk evaluation program, and data structure | |
US9665395B1 (en) | Computerized method of identifying process data exceptions | |
JP2008109101A (ja) | モデル作成装置並びにプロセス異常分析装置およびそれらの方法並びにプログラム | |
CN111796981A (zh) | 信息处理装置、信息处理方法和学习装置 | |
Huang et al. | Design of effective inspection schema for imperfect production systems | |
JP6861653B2 (ja) | 監視制御支援装置 |
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: 20833158 Country of ref document: EP Kind code of ref document: A1 |
|
ENP | Entry into the national phase |
Ref document number: 2021527524 Country of ref document: JP Kind code of ref document: A |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
REG | Reference to national code |
Ref country code: BR Ref legal event code: B01A Ref document number: 112021026482 Country of ref document: BR |
|
ENP | Entry into the national phase |
Ref document number: 2020302438 Country of ref document: AU Date of ref document: 20200529 Kind code of ref document: A |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2020833158 Country of ref document: EP |
|
ENP | Entry into the national phase |
Ref document number: 112021026482 Country of ref document: BR Kind code of ref document: A2 Effective date: 20211227 |
|
WWE | Wipo information: entry into national phase |
Ref document number: 521431213 Country of ref document: SA |