Information processing device, inference device, machine learning device, information processing method, inference method, and machine learning method

Abstract

The present invention provides an information processing device that enables appropriate abnormality diagnosis of equipment supplied with crude oil. [Solution] The information processing device 7 includes a crude oil component data acquisition unit 700 that acquires crude oil component data for prediction, which indicates the components of the crude oil supplied to the equipment to be predicted; an equipment status data acquisition unit 701 that acquires equipment status data for prediction, which indicates the state of the equipment to be predicted; and a prediction processing unit 702 that predicts the occurrence of abnormalities in the equipment to be predicted based on equipment diagnostic data output by inputting the crude oil component data for prediction and the equipment status data for prediction into the learning model 14. The learning model 14 is a trained model that has been trained by machine learning to learn the relationship between the crude oil component data and equipment status data to be learned and equipment diagnostic data indicating the diagnostic result of the equipment to be learned when crude oil with the components indicated by the crude oil component data to be learned is supplied to the equipment to be learned in the state indicated by the equipment status data to be learned.

Description

Technical Field

[0001] The present invention relates to an information processing apparatus, an inference apparatus, a machine learning apparatus, an information processing method, an inference method, and a machine learning method.

Background Art

[0002] Since a plant is composed of a variety of devices, it is necessary to diagnose whether there is an abnormality in each device, and perform maintenance such as inspection, repair, and replacement of the device as needed. For example, Patent Document 1 discloses that condition diagnosis of a device is performed by performing vibration analysis, impact data analysis, lubricating oil analysis, etc. based on data acquired by sensors attached to the devices of a plant. Further, Patent Document 2 discloses that abnormality diagnosis of a device is performed based on data such as plant parameters (output, flow rate, radiation dose, etc.), operation history, repair history, periodic inspection history, material specifications, and structure.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Patent Document 2

Summary of the Invention

Problems to be Solved by the Invention

[0004] In a plant having crude oil in the supply flow, since crude oil is supplied to each device, an abnormality of the device may occur due to components contained in the crude oil. For example, the sulfur component contained in crude oil has a great influence on the deterioration of the device, and depending on the components of the crude oil supplied to each device, the timing and probability of an abnormality occurring in the device are different. Therefore, there was a risk that it was not possible to appropriately perform abnormality diagnosis of a device to which crude oil is supplied only with the data disclosed in Patent Document 1 and Patent Document 2.

[0005] In view of the above problems, the present invention aims to provide an information processing device, an inference device, a machine learning device, an information processing method, an inference method, and a machine learning method that enable appropriate abnormality diagnosis of equipment supplied with crude oil. [Means for solving the problem]

[0006] To achieve the above objective, an information processing apparatus according to one aspect of the present invention is: An information processing device that predicts the occurrence of abnormalities in equipment installed in a plant, A crude oil component data acquisition unit acquires crude oil component data that indicates the composition of the crude oil supplied to the equipment to be predicted, A device status data acquisition unit that acquires device status data indicating the status of the device to be predicted, The system includes a prediction processing unit that predicts the occurrence of abnormalities in the equipment to be predicted based on equipment diagnostic data output by inputting the crude oil component data to be predicted acquired by the crude oil component data acquisition unit and the equipment status data to be predicted acquired by the equipment status data acquisition unit into a learning model, and then performing the prediction. The aforementioned learning model, The relationship between the crude oil component data and the equipment status data to be learned, and the equipment diagnostic data showing the diagnostic result of the equipment to be learned when the crude oil with the components indicated by the crude oil component data to be learned is supplied to the equipment to be learned in the state indicated by the equipment status data to be learned, is determined by the This is a pre-trained model that has been trained using machine learning. [Effects of the Invention]

[0007] According to one aspect of the present invention, an information processing device is used to predict the occurrence of abnormalities in equipment based on equipment diagnostic data output by inputting crude oil component data to be predicted and equipment status data to be predicted into a learning model. Therefore, abnormalities in equipment supplied with crude oil can be appropriately diagnosed.

[0008] Other issues, configurations, and effects will be clarified in the embodiments for carrying out the invention described later. [Brief explanation of the drawing]

[0009] [Figure 1] This is an overall diagram showing an example of plant management system 1 and plant 10. [Figure 2] This is a data configuration diagram showing an example of a crude oil database 50, an equipment condition history database 51, and an equipment diagnostic history database 52. [Figure 3] This is a block diagram showing an example of a machine learning device 6. [Figure 4] This is an explanatory diagram showing an example of training data 13 and training model 14. [Figure 5] This is a block diagram showing an example of an information processing device 7. [Figure 6] This is the first functional diagram showing an example of the information processing device 7. [Figure 7] This is a second functional diagram showing an example of the information processing device 7. [Figure 8] This is a third functional diagram illustrating an example of the information processing device 7. [Figure 9] This is the fourth functional diagram showing an example of the information processing device 7. [Figure 10] This is a hardware configuration diagram showing an example of the Computer 900. [Figure 11] This flowchart shows an example of a machine learning method using the machine learning device 6. [Figure 12] This flowchart shows an example of an information processing method by the information processing device 7. [Modes for carrying out the invention]

[0010] Hereinafter, embodiments for carrying out the present invention will be described with reference to the drawings. In the following, the scope necessary for explaining how to achieve the objectives of the present invention will be schematically shown, and the scope necessary for explaining the relevant parts of the present invention will be mainly explained, with any parts that are omitted from explanation being based on prior art.

[0011] (Plant Management System 1) FIG. 1 is an overall view showing an example of the plant management system 1 and the plant 10. The plant management system 1 according to the present embodiment functions as a system that predicts the occurrence status of abnormalities in the equipment 12 installed in the plant 10 and manages the plant 10 based on the prediction results. The plant 10 has crude oil in the supply flow, and examples include, but are not limited to, oil refining plants, chemical plants, and the like.

[0012] The plant 10 is composed of a plurality of devices 12, and crude oil is supplied from the crude oil tank 11 to each device 12 to perform a predetermined manufacturing process. At that time, a plurality of crude oil tanks 11 are installed, and the crude oil stored in the plurality of crude oil tanks 11 is mixed at a predetermined mixing ratio, and the mixed crude oil is supplied to each device 12.

[0013] The device 12 includes, but is not limited to, rotating equipment, static equipment, heaters, coolers, heat exchangers, package equipment, instruments, controllers, and the like. The rotating equipment includes, for example, pumps, compressors, turbines, blowers, agitators, valves, and the like. The static equipment includes, for example, tower tanks such as distillation towers and reaction tanks, devices such as reforming devices, sulfur removal devices, and salt removal devices, and piping. The instruments include vibration sensors, temperature sensors, pressure sensors, flow sensors, current sensors, and the like.

[0014] As its main configuration, the plant management system 1 includes, as shown in FIG. 1, a plant operation management device 2, an operator terminal device 3, a crude oil analysis device 4, a data management device 5, a machine learning device 6, and an information processing device 7. Each of the devices 2 to 7 is composed of, for example, a general-purpose or dedicated computer (see FIG. 10 described later), and is connected to a wired or wireless network 8 so as to be able to transmit and receive various data to and from each other. Note that the number of each of the devices 2 to 7 and the connection configuration of the network 8 are not limited to the example in FIG. 1 and may be changed as appropriate.

[0015] The plant operation management device 2 is a device that controls the manufacturing process, performs emergency shutdowns, and monitors abnormalities in each piece of equipment 12 by sending and receiving various equipment signals (sensor signals and control signals) to and from multiple pieces of equipment 12 installed at various locations in the plant 10. Specifically, the plant operation management device 2 receives sensor signals (also called input signals) from the equipment 12 that functions as instruments, and transmits control signals (also called output signals) to the equipment 12 that functions as controllers based on the sensor information indicated by those sensor signals. The plant operation management device 2 also transmits equipment status data based on the sensor signals to the data management device 5.

[0016] The worker terminal device 3 is a terminal device used by workers (operators, maintenance workers, analysts, managers, etc.) working at the plant 10, and may be a stationary or portable device. The worker terminal device 3 has programs such as applications and browsers installed on it, accepts various input operations, and outputs various information via a display screen or voice.

[0017] The crude oil analyzer 4 is a device that analyzes crude oil samples supplied from the crude oil tank 11 and taken at predetermined sample locations S1-Si and Sm, and outputs crude oil component data indicating the components of the crude oil as an analysis result. At sample locations S1-Si, crude oil stored in each crude oil tank 11 is taken, while at sample location Sm, mixed crude oil mixed at a predetermined mixing ratio is taken. If the crude oil analyzer 4 has a communication function, the crude oil analyzer 4 transmits the crude oil component data to the data management device 5. If the crude oil analyzer 4 does not have a communication function, the operator performs an input operation to input the analysis results of the crude oil analyzer 4 to the operator terminal device 3, and the operator terminal device 3 transmits the crude oil component data based on that input operation to the data management device 5.

[0018] The data management device 5 is a device that manages various data related to the plant 10 as a database, and includes a crude oil database 50, an equipment status history database 51, and an equipment diagnostic history database 52.

[0019] The crude oil database 50 is a database in which crude oil component data showing the components of crude oil obtained as a result of analysis when a crude oil sample is analyzed by the crude oil analyzer 4 is registered and stored. The equipment status history database 51 is a database in which equipment status data showing the state of equipment 12 obtained as a result of measurement taken at a predetermined measurement cycle by each equipment 12 (mainly instruments) when a manufacturing process is carried out by each equipment 12 is registered and stored. The equipment diagnosis history database 52 is a database in which equipment diagnosis data showing the diagnosis results of equipment 12 is registered and stored when a diagnosis regarding the occurrence of abnormalities in each equipment 12 is performed. Note that each database 50 to 52 may register and store data related to multiple plants 10, and the details of the data structure will be described later.

[0020] The machine learning device 6 is the device that operates as the main component of the machine learning learning phase. For example, the machine learning device 6 acquires training data 13 (details described later) from the data management device 5 and generates a learning model 14 (details described later) to be used by the information processing device 7 using an arbitrary learning algorithm based on the training data 13. The trained learning model 14 is provided to the information processing device 7 via the network 8, recording media, etc.

[0021] The information processing device 7 is a device that operates as the main component in the inference phase of machine learning. The information processing device 7 uses the learning model 14 generated by the machine learning device 6 to predict the occurrence of abnormalities in the target equipment 12. The prediction results of the abnormality occurrence are provided, for example, to the operator terminal device 3 and presented to the operator. Alternatively, the prediction results of the abnormality occurrence may be provided to the data management device 5 and registered in the equipment diagnostic history database 52, etc.

[0022] Figure 2 is a data configuration diagram showing an example of a crude oil database 50, an equipment condition history database 51, and an equipment diagnostic history database 52.

[0023] The crude oil database 50 registers, for example, a crude oil ID to identify the crude oil, the sample date and time, the sample location S1~Si, Sm indicating where the crude oil sample was taken, and crude oil component data indicating the composition of the sample.

[0024] Crude oil component data expresses the composition of crude oil by elemental composition and compound composition. In elemental composition, for example, as shown in Figure 2, the composition ratio of elements is expressed as follows: carbon: X1 (%), hydrogen: X2 (%), sulfur: X3 (%), nitrogen: X4 (%), oxygen: X5 (%), metal: X6 (%). In processed product composition, for example, the composition ratio of compounds is expressed as follows: paraffinic hydrocarbons: Y1 (%), olefinic hydrocarbons: Y2 (%), naphthenic hydrocarbons: Y3 (%), aromatic hydrocarbons: Y4 (%), sulfur compounds: Y5 (%), nitrogen compounds: Y6 (%), oxygen compounds: Y7 (%), metal compounds: Y8 (%). In the example in Figure 2, the composition ratio of each element included in the crude oil component data is indicated as X1 to X6, and the specific values ​​are omitted.

[0025] The equipment status history database 51 registers equipment status data, such as the crude oil ID, measurement date and time, and the status of equipment 12, using the equipment ID to identify equipment 12 as the key. By using the equipment ID to refer to, for example, the drawing data of plant 10 (plot plan, P&ID diagram (Piping & Instrument Diagram), I / O list, etc.), the type, placement relationship, and connection relationship of equipment 12 identified by the equipment ID are identified, as well as the function and role of equipment 12 in the manufacturing process. The crude oil ID identifies the crude oil being supplied to equipment 12 identified by the equipment ID.

[0026] The equipment status data is time-series data showing the change in the state of equipment 12 over time when equipment 12 is in operation. The equipment status data is, for example, time-series data of equipment status parameters recorded by equipment 12, which functions as an instrument, at a predetermined measurement period (seconds, minutes, hours, etc.). In the example in Figure 2, the equipment status parameters included in the equipment status data are registered as follows: vibration: D1, temperature: D2, pressure: D3, flow rate (or flow velocity): D4, and current value: D5. Note that for a single piece of equipment 12, for example, equipment status parameters measured upstream of equipment 12 and equipment status parameters measured downstream of equipment 12 may be registered. Also, the values ​​of the equipment status parameters may be registered as normalized values ​​in the range of 0 to 1, for example. In the example in Figure 2, each equipment status parameter included in the equipment status data is written as D1 to D6, and the specific values ​​are omitted.

[0027] The equipment diagnostic history database 52 registers equipment diagnostic data, such as the crude oil ID, diagnostic date and time, and diagnostic results for equipment 12, using the equipment ID to identify equipment 12 as the key. The crude oil ID identifies the crude oil being supplied to equipment 12 when the equipment 12 identified by the equipment ID is diagnosed.

[0028] The device diagnostic data shows the diagnostic results of device 12, and for example, it may be a record of the diagnostic results of whether or not an abnormality occurs in device 12, or as shown in Figure 2, device 1 In step 2, the diagnostic results for the remaining lifespan until an abnormality occurs may also be recorded. In this case, the diagnostic results for the presence or absence of an abnormality are classified as a binary classification. The diagnostic results for the remaining lifespan are classified as a multi-class classification, and the remaining lifespan period may be changed as appropriate. The diagnosis of equipment 12 may be performed, for example, by the plant operation management device 2 performing abnormality monitoring based on sensor signals, or by maintenance workers performing maintenance work on equipment 12.

[0029] (Machine learning device 6) Figure 3 is a block diagram showing an example of a machine learning device 6. The machine learning device 6 comprises a control unit 60, a communication unit 61, and a storage unit 62.

[0030] The control unit 60 functions as a training data acquisition unit 600 and a machine learning unit 601 by executing a machine learning program 620 stored in the memory unit 62. The communication unit 61 is connected to an external device via the network 8 and functions as a communication interface for sending and receiving various types of data.

[0031] The learning data acquisition unit 600 is connected to external devices (e.g., worker terminal device 3, data management device 5, etc.) via the communication unit 61 and network 8, and acquires learning data 13, which consists of crude oil component data and equipment status data as input data, and equipment diagnostic data as output data. The learning data 13 is used as training data, validation data, and test data in supervised learning. In addition, the output data of the learning data 13 is used as correct data (hereinafter referred to as correct labels) in supervised learning.

[0032] The machine learning unit 601 performs machine learning on the learning model 14 using multiple sets of training data 13 stored in the memory unit 62. When performing machine learning, the machine learning unit 601 can employ any method, such as online learning, batch learning, or mini-batch learning. The machine learning unit 601 may also perform predetermined preprocessing on the input data to be input to the learning model 14, or perform predetermined postprocessing on the output data output from the learning model 14.

[0033] In addition to the machine learning program 620, the memory unit 62 temporarily stores the training data 13 acquired by the training data acquisition unit 600, and also stores the trained learning model 14 (specifically, the adjusted weight parameter set) generated by the machine learning unit 601. The trained learning model 14 stored in the memory unit 62 is provided to the actual system (for example, the information processing device 7) via the network 8 or a recording medium.

[0034] Figure 4 is an explanatory diagram showing an example of training data 13 and a training model 14. The training data 13 used for machine learning of the training model 14 consists of input data and output data.

[0035] The input data includes data on the components of the crude oil supplied to the training device 12, and data on the state of the training device 12.

[0036] The crude oil component data to be learned, included in the input data, indicates the components of the crude oil supplied to the learning device 12 in terms of elemental composition and compound composition. When mixed crude oil is supplied to each device 12, the crude oil component data at sample location Sm is used. When crude oil is supplied to each device 12 from a single crude oil tank 11, the crude oil component data at each of the sample locations S1 to Si may be used. For details of the data, please refer to the Crude Oil Database. Since this is the same as the crude oil component data registered in S50, the explanation will be omitted.

[0037] The device status data included in the input data indicates the state of the device 12 being studied. Since the details of the data are the same as those registered in the device status history database 51, a detailed explanation is omitted.

[0038] The output data includes instrument diagnostic data showing the diagnostic results of the instrument 12 when crude oil with the components indicated by the crude oil component data to be learned is supplied to the instrument 12 in the state indicated by the instrument state data to be learned.

[0039] The device diagnostic data included in the output data shows the diagnostic results of the device 12 being studied. Since the details of the data are the same as those registered in the device diagnostic history database 52, a detailed explanation is omitted.

[0040] In the example in Figure 4, the crude oil component data to be learned represents the components of crude oil by elemental composition, and the equipment status data to be learned is shown as time-series data of equipment status parameters such as vibration, temperature, pressure, flow rate, and current value. Also in the example in Figure 4, the equipment diagnostic data to be learned shows the diagnostic results of the remaining life of equipment 12, and is shown as a classification result categorized as no abnormality, abnormality (no remaining life), remaining life of 1 year, remaining life of 2 years, and remaining life of 3 years.

[0041] The learning data acquisition unit 600 acquires learning data 13 by referring to various data registered in each database 50 to 52 or by receiving input operations from the operator terminal device 3. When the learning data acquisition unit 600 refers to each database 50 to 52, it acquires learning data 13 by associating crude oil component data and equipment status data with equipment diagnostic data at a specific point in time or period. For example, it acquires learning data 13 by aligning the time axis of each data by using equipment ID, crude oil ID, sample date and time, measurement date and time, and diagnostic date and time.

[0042] The learning model 14 employs, for example, a neural network structure and comprises an input layer 140, a hidden layer 141, and an output layer 142. Synapses (not shown) connect each neuron between each layer, and each synapse is associated with a weight. The weight parameters, consisting of the weights of each synapse, are adjusted by machine learning.

[0043] The input layer 140 has a number of neurons corresponding to the crude oil component data and equipment status data as input data, and each value of the crude oil component data and equipment status data is input to each neuron. The output layer 142 has a number of neurons corresponding to the number of classifications of the equipment diagnostic data as output data (5 in the example in Figure 4), and the prediction results (inference results) of the equipment diagnostic data for the crude oil component data and equipment status data are output as output data. When the learning model 14 is composed of a classification model as shown in Figure 4, the equipment diagnostic data is output as scores (accuracies) for each class, normalized to a predetermined range (e.g., 0 to 1). When the learning model 14 is composed of a regression model, the equipment diagnostic data is output as normalized to a predetermined range (e.g., 0 to 1).

[0044] The machine learning unit 601 inputs multiple sets of training data 13 into the learning model 14 and generates a trained learning model 14 by having the learning model 14 learn the relationship between the input data (crude oil component data and equipment status data) and output data (equipment diagnostic data) contained in the training data 13.

[0045] In this embodiment, the data configuration of the training data 13 and the learning model 14 was described as being as shown in Figure 4. However, multiple data configurations with different conditions may be adopted, for example, differences in machine learning methods, input data, and output data. In such cases, the training data acquisition unit 600 acquires multiple types of training data 13 corresponding to the multiple data configurations with different conditions, and the machine learning unit 601 performs machine learning on each learning model using these training data 13.

[0046] (Information Processing Device 7) Figure 5 is a block diagram showing an example of the information processing device 7. Figure 6 is a first functional diagram showing an example of the information processing device 7. Figure 7 is a second functional diagram showing an example of the information processing device 7. Figure 8 is a third functional diagram showing an example of the information processing device 7. Figure 9 is a fourth functional diagram showing an example of the information processing device 7.

[0047] The information processing device 7 comprises a control unit 70, a communication unit 71, and a storage unit 72. The control unit 70 functions as a crude oil component data acquisition unit 700, an equipment status data acquisition unit 701, a prediction processing unit 702, and an output processing unit 703 by executing an information processing program 720 stored in the storage unit 72. The communication unit 71 is connected to an external device via a network 8 and functions as a communication interface for sending and receiving various types of data.

[0048] The crude oil component data acquisition unit 700 is connected to external devices (e.g., worker terminal device 3, crude oil analyzer 4) via the communication unit 71 and network 8, and acquires predicted crude oil component data indicating the components of the crude oil supplied to the prediction target device 12. In the example in Figure 6, for example, crude oil component data acquired at sample location Sm and identified by crude oil ID "Sm-001" is shown.

[0049] As for specific acquisition methods, for example, when an analysis is performed on a sample of crude oil supplied to the equipment 12 to be predicted, the crude oil component acquisition unit 700 acquires the crude oil component data to be predicted by receiving the crude oil component data, which shows the crude oil components obtained as a result of the analysis, from the operator terminal device 3 or the crude oil analyzer 4. In addition, if the analysis results of the crude oil sample supplied to the equipment 12 to be predicted are registered in the crude oil database 50, the crude oil component acquisition unit 700 acquires the crude oil component data to be predicted by referring to the crude oil database 50. Furthermore, the crude oil component acquisition unit 700 may acquire the crude oil component data to be predicted specified in the operation plan by referring to the operation plan of the plant 10 managed by the plant operation management device 2.

[0050] Furthermore, when multiple types of crude oil are mixed at a predetermined mixing ratio and the mixed crude oil is supplied to the prediction target equipment 12, the crude oil component data acquisition unit 700 may acquire prediction target mixed crude oil component data indicating the components of the mixed crude oil based on multiple crude oil component data indicating the components of each of the multiple types of crude oil and the mixing ratio, as shown in Figure 7. For example, the mixed crude oil component data can be obtained by multiplying each component included in the multiple crude oil component data by the mixing ratio and summing them up. In the example in Figure 7, the case where mixed crude oil component data Sm1 is acquired when the crude oil component data obtained at sample location S1 and identified by crude oil ID "S1-011" is "10%", the crude oil component data obtained at sample location S2 and identified by crude oil ID "S2-011" is "20%", ..., and the crude oil component data obtained at sample location Si and identified by crude oil ID "Si-011" is "40%" is mixed at a mixing ratio R1.

[0051] Similarly, when multiple types of crude oil are mixed in a predetermined mixing ratio and the mixed crude oil is supplied to the prediction target device 12, the crude oil component data acquisition unit 700, as shown in Figures 8 and 9, Alternatively, based on multiple types of crude oil component data showing the components of multiple types of crude oil, and multiple mixing ratios, it may be possible to obtain predicted post-mixed crude oil component data showing the components of the mixed crude oil for each mixing ratio. In the examples in Figures 8 and 9, the cases in which post-mixed crude oil component data Sm1, ..., and post-mixed crude oil component data Smj are obtained for each of the multiple mixing ratios R1, ..., and mixing ratio Rj.

[0052] The equipment status data acquisition unit 701 is connected to external devices (for example, the plant operation management device 2, the data management device 5, etc.) via the communication unit 71 and the network 8, and acquires equipment status data that indicates the status of the equipment 12 to be predicted.

[0053] For example, the equipment status data acquisition unit 701 acquires equipment status data for prediction by receiving sensor signals from instruments installed on the equipment 12 to be predicted (in the examples in Figures 6 to 9, this includes a vibration sensor, temperature sensor, pressure sensor, flow sensor, and current sensor). Furthermore, if the measurement results of the instruments installed on the equipment 12 to be predicted are registered in the equipment status history database 51, the equipment status data acquisition unit 701 acquires the equipment status data for prediction by referring to the equipment status history database 51. In addition, the equipment status data acquisition unit 701 may acquire the equipment status data for prediction planned in the operation plan by referring to the operation plan of the plant 10 managed by the plant operation management device 2.

[0054] The prediction processing unit 702 predicts the abnormality status of the equipment 12 to be predicted based on the equipment diagnostic data output by inputting the crude oil component data to be predicted, acquired by the crude oil component data acquisition unit 700, and the equipment status data to be predicted, acquired by the equipment status data acquisition unit 701, into the learning model 14. At that time, the prediction processing unit 702 may perform predetermined pre-processing on the input data (crude oil component data and equipment status data) input to the learning model 14, or may perform predetermined post-processing on the output data (equipment diagnostic data) output from the learning model 14.

[0055] In the example in Figure 6, based on the equipment diagnostic data output by inputting crude oil component data identified by crude oil ID "Sm-001" and equipment status data into the learning model 14, the abnormality status of the equipment 12 to be predicted is shown to be that no abnormalities have occurred and the remaining lifespan is 1 year.

[0056] Furthermore, when multiple types of crude oil are mixed at a predetermined mixing ratio and the mixed crude oil is supplied to the equipment 12 to be predicted, the prediction processing unit 702 may predict the abnormality status of the equipment 12 to be predicted based on equipment diagnostic data output by inputting the mixed crude oil component data of the equipment to be predicted, acquired by the crude oil component data acquisition unit 700, and the equipment status data of the equipment to be predicted, acquired by the equipment status data acquisition unit 701, into the learning model 14. In the example in Figure 7, the case is illustrated in which, based on the equipment diagnostic data output by inputting the mixed crude oil component data Sm1 when mixed at mixing ratio R1 and the equipment status data, it is predicted that no abnormality has occurred in the equipment 12 to be predicted and that the remaining lifespan is 2 years.

[0057] Similarly, when multiple types of crude oil are mixed at predetermined mixing ratios and the mixed crude oil is supplied to the equipment 12 to be predicted, the prediction processing unit 702 may input the mixed crude oil component data acquired for each mixing ratio by the crude oil component data acquisition unit 700 and the equipment status data acquired for each mixing ratio by the equipment status data acquisition unit 701 into the learning model 14, and then predict the occurrence of abnormalities in the equipment 12 to be predicted for each mixing ratio based on the equipment diagnostic data output for each mixing ratio.

[0058] In this case, if the equipment diagnostic data indicates whether or not an abnormality occurs in equipment 12, the prediction processing unit 702 predicts whether or not an abnormality occurs in equipment 12 for each mixing ratio as the abnormality occurrence situation. Also, if the equipment diagnostic data indicates the diagnostic result of the remaining lifespan until an abnormality occurs in equipment 12, the prediction processing unit 702 predicts the remaining lifespan of equipment 12 for each mixing ratio as the abnormality occurrence situation. In the example in Figure 8, the mixed crude oil component data Sm1, ..., Smj and equipment status data for mixing at multiple mixing ratios R1, ..., Rj are input to the learning model 14, respectively, and the equipment diagnostic data E1, ..., Ej are output. The figure shows the case where, as the abnormality occurrence situation for equipment 12, the remaining lifespan is predicted to be 1 year for mixing ratio R1 and 3 years for mixing ratio Rj. Furthermore, the predicted remaining lifespan of the equipment 12 for each mixing ratio can also be represented as a scatter plot with the mixing ratio on the horizontal axis and the remaining lifespan on the vertical axis, for example, as shown in Figure 8.

[0059] Furthermore, if the equipment diagnostic data indicates the remaining lifespan until an abnormality occurs in equipment 12, the prediction processing unit 702 predicts the remaining lifespan of the equipment 12 for each mixing ratio as an abnormality occurrence situation, and identifies the mixing ratio among multiple mixing ratios that satisfies the predetermined lifespan condition for the equipment 12 being predicted. In the example in Figure 9, similar to Figure 8, the diagram shows a case where, as an abnormality occurrence situation for the equipment 12 being predicted, the remaining lifespan is predicted to be 1 year for mixing ratio R1 and 3 years for mixing ratio Rj. The diagram then shows a case where mixing ratios R3, ..., Rj are identified as mixing ratios that satisfy the predetermined lifespan condition of a remaining lifespan of 3 years. The lifespan condition is stored in the storage unit 72 as data that can be displayed and edited on, for example, the operator terminal device 3.

[0060] The output processing unit 703 performs output processing to output the equipment diagnostic data output from the learning model 14 and the predicted abnormality status of the equipment 12 based on the equipment diagnostic data. For example, the output processing unit 703 may transmit screen information for displaying the equipment diagnostic data and the abnormality status of the equipment 12 to the operator terminal device 3 so that the information is displayed on the display screen of the operator terminal device 3, or it may transmit the equipment diagnostic data and the abnormality status of the equipment 12 to the data management device 5 so that the information is registered in the equipment diagnostic history database 52.

[0061] The output processing unit 703 may also output maintenance information based on equipment diagnostic data as part of its output processing. For example, the storage unit 72 or data management device 5 may store the maintenance deadline for the remaining lifespan (data that can be displayed and edited on the operator terminal device 3), and the output processing unit 703 may output maintenance information when it determines that the equipment diagnostic data is below the maintenance deadline.

[0062] The memory unit 72 stores the information processing program 720 as well as the trained learning models 14 used by the prediction processing unit 702. The number of learning models 14 stored in the memory unit 72 is not limited to the above example; for example, multiple trained models with different conditions, such as differences in machine learning methods, input data, and output data, may be stored and used selectively or in parallel. The memory unit 72 may also be replaced by the memory unit of an external computer (which may be the data management device 5 in this embodiment), in which case the prediction processing unit 702 only needs to access the external computer.

[0063] (Hardware configuration of each device) Figure 10 is a hardware configuration diagram showing an example of the computer 900 that constitutes each device. Each device 2 to 7 in the plant management system 1 is composed of a general-purpose or dedicated computer 900.

[0064] As shown in Figure 10, the computer 900 comprises, as its main components, a bus 910, a processor 912, memory 914, an input device 916, an output device 917, a display device 918, a storage device 920, a communication interface unit 922, an external device interface unit 924, an I / O device interface unit 926, and a media input / output unit 928. Note that the above components may be omitted as appropriate depending on the intended use of the computer 900.

[0065] The processor 912 consists of one or more arithmetic processing units (CPU (Central Processing Unit), MPU (Micro-Processing Unit), DSP (Digital Signal Processor), GPU (Graphics Processing Unit), NPU (Neural Processing Unit), etc.) and operates as a control unit that oversees the entire computer 900. The memory 914 stores various data and programs 930 and consists of volatile memory (DRAM, SRAM, etc.) that functions as main memory, and non-volatile memory (ROM), flash memory, etc.

[0066] The input device 916 consists of, for example, a keyboard, mouse, numeric keypad, or electronic pen, and functions as an input unit. The output device 917 consists of, for example, a sound (voice) output device or a vibration device, and functions as an output unit. The display device 918 consists of, for example, a liquid crystal display, an organic EL display, electronic paper, or a projector, and functions as an output unit. The input device 916 and the display device 918 may be configured as an integrated unit, such as a touch panel display. The storage device 920 consists of, for example, an HDD or SSD, and functions as a storage unit. The storage device 920 stores various data necessary for the execution of the operating system and program 930.

[0067] The communication I / F unit 922 is connected by wire or wireless to a network 940 such as the Internet or an intranet (which may be the same as network 8 in Figure 1) and functions as a communication unit that sends and receives data with other computers according to a predetermined communication standard. The external device I / F unit 924 is connected by wire or wireless to external devices 950 such as cameras, printers, scanners, and reader / writers and functions as a communication unit that sends and receives data with external devices 950 according to a predetermined communication standard. The I / O device I / F unit 926 is connected to I / O devices 960 such as various sensors and actuators and functions as a communication unit that sends and receives various signals and data with the I / O devices 960, such as detection signals from sensors and control signals to actuators. The media input / output unit 928 consists of, for example, a drive device such as a DVD drive or CD drive, a memory card slot, and a USB connector, and reads and writes data to media (non-temporary storage media) 970 such as DVDs, CDs, memory cards, and USB memory.

[0068] In the computer 900 having the above configuration, the processor 912 calls and executes the program 930 stored in the storage device 920 in the memory 914, and controls various parts of the computer 900 via the bus 910. The program 930 may also be stored in the memory 914 instead of the storage device 920. The program 930 may be recorded on the media 970 in an installable or executable file format and provided to the computer 900 via the media input / output unit 928. The program 930 may also be provided to the computer 900 by downloading it via the network 940 through the communication interface unit 922. Furthermore, the computer 900 may implement various functions realized by the processor 912 executing the program 930 using hardware such as an FPGA (Field-Programmable Gate Array) or ASIC (Application Specific Integrated Circuit).

[0069] Computer 900 is an electronic device of any form, consisting of, for example, a stationary computer or a portable computer. Computer 900 may be a client computer, a server computer, a cloud computer, or an embedded computer such as a control panel or controller (including microcontrollers, programmable logic controllers, and sequencers).

[0070] (Operation of Plant Management System 1) The operation of Plant Management System 1 will be described below.

[0071] (Machine learning methods) Figure 11 is a flowchart illustrating an example of a machine learning method using the machine learning device 6. In the following explanation, it is assumed that each database 50-52 contains a large amount of data.

[0072] First, in step S100, when the training data acquisition unit 600 receives, for example, operation information from the operator terminal device 3 instructing it to start machine learning, it refers to each database 50 to 52, acquires a desired number of training data 13, and temporarily stores the acquired training data 13 in the storage unit 62.

[0073] Next, in step S110, the machine learning unit 601 prepares a pre-training model 14 in order to start machine learning. The pre-training model 14 prepared here is, for example, a neural network model, in which the weights of each synapse are set to initial values.

[0074] Next, in step S120, the machine learning unit 601 randomly selects, for example, one set of training data 13 from the multiple sets of training data 13 stored in the memory unit 62.

[0075] Next, in step S130, the machine learning unit 601 inputs the crude oil component data and equipment status data (input data) contained in a set of training data 13 to the input layer 140 of the prepared pre-training (or training) learning model 14. As a result, equipment diagnostic data (output data) is output as an inference result from the output layer 142 of the learning model 14, but this output data is generated by the pre-training (or training) learning model 14. Therefore, in the pre-training (or training) state, the output data output as an inference result shows different information from the equipment diagnostic data (ground truth labels) contained in the training data 13.

[0076] Next, in step S140, the machine learning unit 601 compares the equipment diagnostic data (ground truth labels) included in the set of training data 13 acquired in step S120 with the equipment diagnostic data (output data) output as an inference result from the output layer 142 in step S130, and performs machine learning by adjusting the weight of each synapse (backpropagation). In this way, the machine learning unit 601 trains the learning model 14 on the relationship between crude oil component data, equipment status data and equipment diagnostic data.

[0077] Next, in step S150, the machine learning unit 601 determines whether predetermined learning termination conditions have been met, for example, based on the evaluation value of the error function, which is based on the device diagnostic data (correct labels) included in the training data 13 and the device diagnostic data output as an inference result (output data), or based on the remaining number of untrained training data 13 stored in the storage unit 62.

[0078] If, in step S150, the machine learning unit 601 determines that the learning termination condition has not been met and decides to continue machine learning (No in step S150), it returns to step S120. Next, steps S120 to S140 are performed multiple times on the learning model 14 that is currently being trained, using the untrained training data 13. On the other hand, in step S150, if the machine learning unit 601 determines that the learning termination condition has been met and that machine learning should be terminated (Yes in step S150), the process proceeds to step S160.

[0079] Then, in step S160, the machine learning unit 601 stores the trained model 14 (set weight parameters) generated by adjusting the weights associated with each synapse in the storage unit 62, and the series of machine learning methods shown in Figure 11 is completed. In the machine learning method, step S100 corresponds to the training data acquisition step, steps S110 to S150 are the machine learning steps, and step S160 is the trained model storage step.

[0080] As described above, the machine learning apparatus 6 and machine learning method according to this embodiment provide a learning model 14 capable of inferring output data, including equipment diagnostic data, from input data, including crude oil component data and equipment status data.

[0081] (Information processing methods) Figure 12 is a flowchart showing an example of an information processing method by the information processing device 7. Below, we will describe an example of operation when an operator uses the operator terminal device 3 to predict the occurrence of an abnormality in the equipment 12 to be predicted. Furthermore, it will be assumed that the storage unit 72 of the information processing device 7 has a trained learning model 14 stored in it.

[0082] First, in step S200, when the operator terminal device 3 receives prediction conditions, such as those specified by the operator on the display screen, specifying the crude oil to be predicted (for example, by crude oil ID or mixing ratio) and the equipment 12 to be predicted (for example, by equipment ID), it transmits these prediction conditions to the information processing device 7. Note that multiple crude oils may be specified as the crude oil to be predicted, and multiple pieces of equipment 12 may be specified as the equipment 12 to be predicted.

[0083] Next, in step S210, when the crude oil component data acquisition unit 700 of the information processing device 7 receives the prediction conditions transmitted in step S200, it acquires the petroleum component data to be predicted according to those prediction conditions.

[0084] For example, the crude oil component data acquisition unit 700 obtains the crude oil component data to be predicted by referring to the crude oil database 50 using the crude oil ID as a key, as shown in Figure 6. Alternatively, the crude oil component data acquisition unit 700 may obtain the mixed crude oil component data to be predicted based on the crude oil component data obtained using the crude oil ID as a key and the mixing ratio, as shown in Figure 7, or it may obtain the mixed crude oil component data to be predicted for each mixing ratio, as shown in Figures 8 and 9.

[0085] Next, in step S220, the equipment status data acquisition unit 701 acquires the equipment status data to be predicted according to the prediction conditions transmitted in step S200. For example, the equipment status data acquisition unit 701 acquires the equipment status data to be predicted by referring to the equipment status history database 51 using the equipment ID as the key.

[0086] Next, in step S230, the prediction processing unit 702 inputs the equipment status data to be predicted acquired in step S210 and the crude oil component data to be predicted acquired in step S220 into the learning model 14, and based on the equipment diagnostic data output, predicts the abnormality status of the equipment 12 to be predicted.

[0087] For example, as shown in Figure 6, the crude oil component data acquisition unit 700 acquires the crude oil component data to be predicted. When data is acquired, the learning model 14 receives the data on the target equipment status along with the data on the crude oil components to be predicted, thereby predicting the occurrence of abnormalities in the target equipment 12. Furthermore, when the crude oil component data acquisition unit 700 acquires the data on the mixed crude oil components to be predicted, as shown in Figure 7, the learning model 14 receives the data on the mixed crude oil components to be predicted along with the data on the target equipment status, thereby predicting the occurrence of abnormalities in the target equipment 12. Also, when the crude oil component data acquisition unit 700 acquires the data on the mixed crude oil components to be predicted for each mixing ratio, as shown in Figures 8 and 9, the learning model 14 receives the data on the mixed crude oil components to be predicted, as well as the data on the target equipment status, thereby predicting the occurrence of abnormalities in the target equipment 12 for each mixing ratio.

[0088] Next, in step S240, the output processing unit 703 transmits screen information, including the abnormality status of the equipment 12 predicted in step S230, to the operator terminal device 3 as output processing.

[0089] Then, in step S250, when the worker terminal device 3 receives the screen information transmitted in step S200, the abnormality status of the equipment 12 to be predicted is presented to the worker based on that screen information. As shown in Figures 8 and 9, if the abnormality status of the equipment 12 to be predicted is predicted for each mixing ratio, the abnormality status of the equipment 12 to be predicted for each mixing ratio is presented to the worker.

[0090] In the above series of information processing methods, step S210 corresponds to the crude oil component data acquisition step, step S220 to the equipment status data acquisition step, step S230 to the prediction processing step, and step S240 to the output processing step.

[0091] As described above, according to the information processing device 7 and information processing method of this embodiment, the abnormality status of the equipment 12 to be predicted is predicted based on equipment diagnostic data output by inputting the crude oil component data to be predicted and the equipment status data to be predicted into the learning model 14. Therefore, the abnormality status of the equipment 12 is predicted while taking into account the components contained in the crude oil, for example, the amount of sulfur component contained in the crude oil, so that abnormality diagnosis of the equipment 12 to which the crude oil is supplied can be performed appropriately.

[0092] (Other embodiments) The present invention is not limited to the embodiments described above, and can be implemented with various modifications without departing from the spirit of the invention. All such modifications are included in the technical concept of the present invention.

[0093] In the above embodiment, the plant operation management device 2, data management device 5, machine learning device 6, and information processing device 7 were described as being composed of separate devices. However, these four devices may be composed of a single device, or any two or three of these four devices may be composed of a single device. Furthermore, at least one of the machine learning device 6 and the information processing device 7 may be incorporated into the plant operation management device 2, the worker terminal device 3, or the data management device 5.

[0094] In the above embodiment, a case in which a neural network is used as the learning model 14 that realizes machine learning by the machine learning unit 601 was described, but other machine learning models may also be used. Other machine learning models include, for example, tree-type models such as decision trees and regression trees, ensemble learning such as bagging, boosting, and XGBoost, recurrent neural networks, convolutional neural networks, and neural network-type models such as LSTM (decode). (including deep learning), hierarchical clustering, non-hierarchical clustering, k-nearest neighbors Clustering methods such as k-means method, multivariate methods such as principal component analysis, factor analysis, and logistic regression. Examples include analysis and support vector machines.

[0095] In the above embodiment, the case was described in which the information processing device 7 performs a prediction process to predict the abnormality status of the equipment 12 to be predicted when it receives prediction conditions from the worker terminal device 3. However, even if the information processing device 7 does not receive prediction conditions from the worker terminal device 3, it may still perform the above prediction process and present the prediction results to the worker, for example, if predetermined prediction execution conditions are met. The prediction execution conditions may be specified, for example, as a specific day of the week (every Monday) or a specific date (the last day of each month).

[0096] (Inference device, inference method, and inference program) The present invention can be provided not only in the form of the information processing apparatus 7 (information processing method or information processing program 720) according to the above embodiment, but also in the form of an inference device (inference method or inference program) used to infer crude oil component data. In that case, the inference device (inference method or inference program) may include a memory and a processor, the processor of which may execute a series of processes. The series of processes includes a crude oil component data acquisition process (crude oil component data acquisition step) for acquiring predicted crude oil component data indicating the components of crude oil supplied to the equipment 12 to be predicted, an equipment status data acquisition process (equipment status data acquisition step) for acquiring predicted equipment status data indicating the state of the equipment 12 to be predicted, and an inference process (inference step) for inferring a diagnostic result of the equipment 12 to be predicted after acquiring the predicted crude oil component data by the crude oil component data acquisition process and the predicted equipment status data by the equipment status data acquisition process.

[0097] By providing the information processing device (inference method or inference program) in the form of an inference device, it becomes easier to apply to various devices compared to implementing the information processing device 7. It will be obvious to those skilled in the art that when the inference device (inference method or inference program) infers crude oil component data, it may apply the inference method performed by the generation processing unit using the trained learning model generated by the machine learning device and machine learning method according to the above embodiment. [Explanation of symbols]

[0098] 1...Plant management system, 2...Plant operation management device, 3...Worker terminal device, 4...Crude oil analyzer, 5...Data management device, 6...Machine learning device, 7...Information processing device 10...Plant, 11...Crude oil tank, 12...Equipment, 13...Training data 14...Learning model, 50...Crude oil database, 51...Equipment status history database, 52...Equipment diagnostic history database, 60...Control unit, 61...Communication unit, 62...Storage unit, 70...Control unit, 71...Communication unit, 72...Storage unit, 600...Training data acquisition unit, 601...Machine learning unit, 620...Machine learning program, 700...Crude oil component data acquisition unit, 701...Equipment status data acquisition unit, 702...Prediction processing unit, 703...Output processing unit, 720...Information processing program, 900... Computer

Claims

1. An information processing device that predicts the occurrence of abnormalities in equipment installed in a plant, A crude oil component data acquisition unit acquires crude oil component data that indicates the composition of the crude oil supplied to the equipment to be predicted, A device status data acquisition unit that acquires device status data indicating the status of the device to be predicted, The system includes a prediction processing unit that predicts the occurrence of abnormalities in the equipment to be predicted based on equipment diagnostic data output by inputting the crude oil component data to be predicted acquired by the crude oil component data acquisition unit and the equipment status data to be predicted acquired by the equipment status data acquisition unit into a learning model, and then performing the prediction. The aforementioned learning model, This is a trained model that has been trained by machine learning to learn the relationship between the crude oil component data and the equipment status data to be learned, and equipment diagnostic data showing the diagnostic result of the equipment to be learned when the crude oil with the components indicated by the crude oil component data to be learned is supplied to the equipment to be learned in the state indicated by the equipment status data to be learned. Information processing device.

2. The aforementioned equipment status data is This is time-series data showing the change in the state of the equipment over time when the equipment is in operation. The information processing apparatus according to claim 1.

3. The aforementioned crude oil component data acquisition unit is: When multiple types of crude oil are mixed in a predetermined mixing ratio and the mixed crude oil is supplied to the equipment to be predicted, based on multiple crude oil component data representing the components of each of the multiple types of crude oil and the mixing ratio, the equipment acquires mixed crude oil component data representing the components of the mixed crude oil to be predicted. The prediction processing unit, Based on the equipment diagnostic data output by inputting the mixed crude oil component data of the target to be predicted, acquired by the crude oil component data acquisition unit, and the equipment status data of the target to be predicted, into the learning model, the abnormality status of the target equipment is predicted. The information processing apparatus according to claim 1.

4. The aforementioned equipment diagnostic data is This indicates the diagnostic result of whether or not an abnormality occurs in the aforementioned equipment. The prediction processing unit, The aforementioned abnormality occurrence situation is a prediction of whether or not an abnormality will occur in the equipment to be predicted. The information processing apparatus according to claim 3.

5. The aforementioned equipment diagnostic data is This shows the diagnostic results for the remaining lifespan until an abnormality occurs in the aforementioned equipment. The prediction processing unit, As the aforementioned abnormality occurrence situation, predict the remaining lifespan of the equipment to be predicted. The information processing apparatus according to claim 3.

6. The aforementioned crude oil component data acquisition unit is: When multiple types of crude oil are mixed in a predetermined mixing ratio and the mixed crude oil is supplied to the equipment to be predicted, the multiple types of crude oil components each represent the components of the multiple types of crude oil. Based on the data and the multiple mixing ratios, predictable post-mixing crude oil component data, which indicates the components of the crude oil after mixing, is obtained for each mixing ratio. The prediction processing unit, By inputting the mixed crude oil component data of the target to be predicted, acquired for each mixing ratio by the crude oil component data acquisition unit, and the equipment status data of the target to be predicted, acquired by the equipment status data acquisition unit, into the learning model, the abnormality occurrence status of the target equipment is predicted for each mixing ratio based on the equipment diagnostic data output for each mixing ratio. The information processing apparatus according to claim 1.

7. The aforementioned equipment diagnostic data is This indicates the diagnostic result of whether or not an abnormality occurs in the aforementioned equipment. The prediction processing unit, As for the abnormality occurrence situation, the prediction is made for each mixing ratio whether or not an abnormality will occur in the equipment to be predicted. The information processing apparatus according to claim 6.

8. The aforementioned equipment diagnostic data is This shows the diagnostic results for the remaining lifespan until an abnormality occurs in the aforementioned equipment. The prediction processing unit, As for the abnormal occurrence conditions, the remaining lifespan of the equipment to be predicted is predicted for each of the mixing ratios. The information processing apparatus according to claim 6.

9. The aforementioned equipment diagnostic data is This shows the diagnostic results for the remaining lifespan until an abnormality occurs in the aforementioned equipment. The prediction processing unit, As for the abnormal occurrence conditions, the remaining lifespan of the equipment to be predicted is predicted for each mixing ratio, Among the multiple mixing ratios, identify the mixing ratio in which the remaining lifespan of the equipment to be predicted satisfies predetermined lifespan conditions. The information processing apparatus according to claim 6.

10. An inference device comprising memory and a processor, The aforementioned processor, A crude oil component data acquisition process that acquires crude oil component data indicating the composition of the crude oil supplied to the equipment to be predicted, A device status data acquisition process that acquires device status data indicating the status of the device to be predicted, The crude oil component data to be predicted is obtained through the crude oil component data acquisition process, and the equipment status data to be predicted is obtained through the equipment status data acquisition process. Then, an inference process is performed to infer the diagnostic result of the equipment to be predicted. Reasoning device.

11. A training data acquisition unit that acquires multiple sets of training data consisting of input data and output data, A machine learning unit uses multiple sets of the training data acquired by the training data acquisition unit to train a learning model on the relationship between the input data and the output data through machine learning. The machine learning unit has a memory unit that stores the learning model in which the relationship has been learned, The aforementioned input data is The training data shows the composition of the crude oil supplied to the training equipment, This includes learning target device status data that indicates the state of the device to be learned, The output data mentioned above is: The data includes instrument diagnostic data indicating the diagnostic result of the instrument when the crude oil with the components indicated in the crude oil component data to be learned is supplied to the instrument to be learned in the state indicated in the instrument state data to be learned, Machine learning device.

12. An information processing method that uses a computer to predict the occurrence of abnormalities in equipment installed in a plant, A crude oil component data acquisition process that acquires crude oil component data indicating the composition of the crude oil supplied to the equipment to be predicted, A device status data acquisition step, which acquires device status data indicating the state of the device to be predicted, The system includes a prediction processing step which predicts the abnormality status of the equipment to be predicted based on equipment diagnostic data output by inputting the crude oil component data to be predicted obtained by the crude oil component data acquisition step and the equipment status data to be predicted obtained by the equipment status data acquisition step into a learning model, The aforementioned learning model, This is a trained model that has been trained by machine learning to learn the relationship between the crude oil component data and the equipment status data to be learned, and equipment diagnostic data showing the diagnostic result of the equipment to be learned when the crude oil with the components indicated by the crude oil component data to be learned is supplied to the equipment to be learned in the state indicated by the equipment status data to be learned. Information processing methods.

13. An inference method performed by an inference device comprising memory and a processor, The aforementioned processor, A crude oil component data acquisition process that acquires crude oil component data indicating the composition of the crude oil supplied to the equipment to be predicted, A device status data acquisition process that acquires device status data indicating the status of the device to be predicted, The crude oil component data to be predicted is obtained through the crude oil component data acquisition process, and the equipment status data to be predicted is obtained through the equipment status data acquisition process. Then, an inference process is performed to infer the diagnostic result of the equipment to be predicted. Reasoning method.

14. A machine learning method performed by a computer, The training data acquisition process involves acquiring multiple sets of training data consisting of input data and output data, A machine learning step in which a learning model learns the relationship between the input data and the output data using machine learning, using multiple sets of the training data acquired in the training data acquisition step, The system includes a learned model storage step, in which the learned model, which has learned the relationship through the machine learning step, is stored in a memory unit. The aforementioned input data is The training data shows the composition of the crude oil supplied to the training equipment, This includes learning target device status data that indicates the state of the device to be learned, The output data mentioned above is: The data includes instrument diagnostic data indicating the diagnostic result of the instrument when the crude oil with the components indicated in the crude oil component data to be learned is supplied to the instrument to be learned in the state indicated in the instrument state data to be learned, Machine learning methods.

Images