Plant management device and plant management method

Abstract

To predict results of processing performed by a chemical plant by reflecting characteristics of unit operation.SOLUTION: A chemical plant management device includes: a storage part which stores information of a plurality of different unit operations included in processing processes in a chemical plant, log data information related to the processing of the unit operations, and condition information related to a condition in each of the unit operations; an input part which inputs sensor information acquired from a sensor device related to a first unit operation in the chemical plant; and a calculation part which creates a prediction model for predicting results of the processing in the first unit operation based on the log data information and the condition information related to the first unit operation, and stores the prediction model into the storage part in association with the first unit operation.SELECTED DRAWING: Figure 4

Description

[Technical Field]

[0001] The present invention relates to a plant management device and a plant management method. [Background technology]

[0002] Conventionally, there is a technology for creating a model for predicting the quality of a plant's product using AI (Artificial Intelligence) etc. For example, Patent Document 1 describes that "a prediction model creation system includes: a process data processing unit that performs predetermined processing on process data obtained from a chemical plant; and a prediction model creation unit that creates a prediction model that learns characteristics of the process data obtained from the chemical plant based on causal relationship information that defines a combination of first process data as an explanatory variable and second process data or a value corresponding to the second process data as a target variable, among the process data obtained from the chemical plant or the process data processed by the process data processing unit, and the process data processing unit uses the process data to determine a value corresponding to a reaction rate of a process target in a predetermined period." [Prior art documents] [Patent documents]

[0003] [Patent Document 1] WO2021 / 157667 Summary of the Invention [Problem to be solved by the invention]

[0004] In order to analyze sensor information related to plant operation to understand or predict the situation within the plant, model analysis using AI may be performed based on log data within the plant.

[0005] On the other hand, within a plant, there are various processes and unit operations within each process, and the sensor information to be analyzed is information for one of the unit operations. Furthermore, each unit operation within a plant may have its own unique operating conditions and methods.

[0006] However, the inventors have found that, because AI is generally data-driven and lacks the concept of constraints on processes and unit operations or the concept of special operating states that differ from normal operating states, when performing model analysis using AI, the operating conditions and methods specific to the unit operations described above are not taken into account, which can affect analytical efficiency and accuracy. For example, calculations can diverge due to the inclusion of data unrelated to the sensor information of the analysis target, making it impossible to perform efficient analysis, or the basic theory of each unit operation can be ignored, resulting in reduced analytical accuracy. Therefore, a system that can create a model for analyzing sensor information of an analysis target while taking into account constraints such as analysis policies and analysis conditions specific to unit operations is desired. [Means for solving the problem]

[0007] In order to achieve the above-mentioned object, one representative plant management device of the present invention comprises a memory unit that stores information on multiple different unit operations included in a processing process in a chemical plant, log data information related to the processing of the unit operations, and condition information related to the conditions for each of the unit operations; an input unit that inputs sensor information obtained from a sensor device related to a first unit operation in the chemical plant; and a calculation unit that creates a predictive model for predicting the results of processing in the first unit operation based on the sensor information, the log data information related to the first unit operation, and the condition information, and stores the predictive model in the memory unit in association with the first unit operation. Furthermore, one representative plant management method of the present invention includes the steps of inputting sensor information acquired from a sensor device related to a first unit operation included in a plurality of different unit operations included in a processing step in a chemical plant, creating a predictive model for predicting the results of processing in the first unit operation based on the sensor information, log data information related to the processing of the first unit operation, and condition information which is information regarding the conditions in the first unit operation, and storing the predictive model in a specified storage device in association with the first unit operation. [Effects of the Invention]

[0008] According to the present invention, it is possible to predict the results of processing in a chemical plant, reflecting the characteristics of the unit operations. Objects, configurations, and effects other than those described above will become apparent from the following description of the embodiments. [Brief explanation of the drawings]

[0009] [Figure 1] FIG. 1 is an explanatory diagram of reaction prediction in a chemical plant according to an embodiment. [Figure 2] Illustrative diagram of a specific example of reaction prediction [Figure 3] Diagram showing the difference between having and not having condition information [Figure 4] A diagram showing the configuration of a plant management device. [Figure 5] Explanation of plant management device operation [Figure 6] An explanatory diagram of specific examples of grouping [Figure 7] Flowchart showing what happens when you create a predictive model [Figure 8] Flowchart showing the operation of a predictive model [Figure 9] Diagram of automatic plant operation [Figure 10] An explanatory diagram of the case where a batch method is adopted DETAILED DESCRIPTION OF THE INVENTION

[0010] Hereinafter, a representative embodiment for carrying out the present invention will be described with reference to the drawings as appropriate. FIG. 1 is an explanatory diagram of a reaction prediction for a chemical plant according to an embodiment. The plant management device disclosed in the embodiment predicts the results of a process performed using, for example, a reactor. In FIG. 1, sensor data is acquired using sensors regarding the raw materials fed into the reactor and the state of the reactor.

[0011] Here, we will explain unit operations. A unit operation is a procedural element that is composed of algorithms required for initiating, organizing, and controlling phases and defines an independent process. For example, reaction, fractional distillation, mixing, heat transfer, separation, etc. can be treated as unit operations. A plant process includes one or more of these unit operations. An example of a process may be, for example, a production process of a specific product or a part of that production process. The classification of unit operations and processes may be determined depending on the equipment configuration and operation method of the chemical plant, and the specific processing content.

[0012] The disclosed plant management device predicts the results of processing using condition information of unit operations in addition to sensor data. The condition information indicates constraints such as analysis policies and analysis conditions specific to the unit operation, and includes the type of unit operation, a predetermined mathematical model, the type of process, and operating conditions. Examples of unit operation types include reaction and fractional distillation. Examples of mathematical models include reaction rate equations and heat transfer calculation equations. Examples of process types include the type of product produced in the process. Operation conditions include conditions related to temperature, flow rate, etc. Examples include thresholds such as maximum, minimum, and standard deviation, as well as specific values, for sensor data such as temperature and flow rate.

[0013] The plant management device generates a dataset to be provided to the prediction model by adding data, deleting sensor data, correcting sensor data, etc. based on the condition information. Here, the type and period of sensor data required for analysis differ for each unit operation, so by generating a dataset to be provided to the prediction model based on the condition information, only the necessary data is used, making it possible to perform more accurate and efficient predictions. The prediction model then uses the provided dataset to predict the results of the reactor processing (e.g., the quality of the product produced by the reactor). In this way, by adding, deleting, or correcting data based on condition information from the data group obtained from the sensor, it becomes possible to predict reactions while taking into account constraints such as analytical policies and analytical conditions specific to unit operations.

[0014] Figure 2 is an explanatory diagram of a specific example of reaction prediction. For example, when predicting the quality of product A, the plant management system uses, as explanatory variables, the condition information of the unit operations used in the production process of product A in addition to the operational data of product A. This allows for more accurate model creation and more accurate predictions than when performing data analysis using AI using only the operational data of product A as explanatory variables.

[0015] Log data from past productions of product A may be used as the operation data for product A. Reaction kinetics and reaction rate for product A may be used as the condition information.

[0016] Similarly, the quality of products B and C can be predicted by using condition information as explanatory variables in addition to the operation data.

[0017] This section explains how to create a predictive model for use in AI data analysis. The plant management device creates a predictive model for predicting the results of processing in a unit operation based on sensor data, log data related to the unit operation, and condition information. The created predictive model is then associated with the unit operation and stored in a specified memory unit.

[0018] For example, in a chemical plant that produces a single product, a large amount of operational data for a specific unit operation is available, and a predictive model can be created from that operational data. On the other hand, in a plant that handles a wide variety of products, if one tries to analyze the information of each unit operation for each product and each part number, the data will become complex and enormous, reducing the efficiency of the analysis and possibly resulting in a lack of necessary data.

[0019] Therefore, the disclosed plant management device groups unit operations with similar characteristics for the unit operations to be analyzed, enabling analysis of the unit operations corresponding to the sensor information to be analyzed based on similar unit operation conditions and pre-created models. Here, the reason for using similar unit operation conditions and pre-created models is explained. The analytical policies and analytical conditions associated with each unit operation may include common or similar analytical policies and analytical conditions depending on the type of unit operation, the equipment configuration of the chemical plant, and the properties of the target raw materials, intermediates, and products. Therefore, even condition information and log data related to unit operations managed as different unit operations from the unit operation to be analyzed may be used to create a predictive model or a dataset for prediction for the unit operation to be analyzed. For example, unit operations of the same type may include common or similar analytical policies and analytical conditions, even if the unit operations are located at different locations or in different processes. Furthermore, even unit operations of different types may include common or similar analytical policies and analytical conditions depending on the type of feature important for the predictive model or analysis, and the equipment configuration. Furthermore, each unit operation may contain common or similar analytical policies and analytical conditions depending on the nature of the material being processed.

[0020] Specifically, the plant management device creates multiple property groups, each containing one or more unit operations with a predetermined range of properties, based on the properties of the unit operations. Then, among the multiple property groups, a property group containing a unit operation related to the input sensor information is identified, and a predictive model for predicting the processing results of the unit operation is created based on the sensor information and log data information and condition information related to one or more unit operations included in the property group. As another method for creating a model, if the unit operation related to the input sensor information is a first unit operation, a predictive model for the first unit operation may be created using multiple learning models already created for other unit operations included in the property group containing the first unit operation. For example, a predictive model for the first unit operation can be created by ensemble learning of multiple existing learning models.

[0021] In Figure 2, product A and product B are grouped in the same group Gr1. Note that only product C belongs to group Gr2, but in this way, only one unit operation may belong to one group. Grouping can be performed using any characteristics depending on the purpose of the analysis, such as the type of unit operation, the type of process, information on the equipment configuration in the chemical plant, reaction characteristics, information identifying whether it is a batch or continuous process, information on the base identifying the location of the chemical plant, and optimal features for the operating conditions of the unit operation. Furthermore, while the grouping has been described above as creating a set of groups for multiple unit operations, groups may also be created for each of multiple characteristic types related to the unit operations. For example, when focusing on two characteristic types A and B, grouping for characteristic type A (pattern 1) and grouping for characteristic type B (pattern 2) are performed separately. In this case, when sensor data to be analyzed is received, similar groups are identified for each type, and a prediction model or a dataset to be input to the prediction model may be created based on the condition information, log data, or pre-created prediction model related to the unit operations included in the identified similar groups. In this case, the grouping pattern to be referenced may be selected depending on the purpose of the analysis.

[0022] Figure 3 is an explanatory diagram of the difference between the presence and absence of condition information. When using only sensor data, the region selected by the AI ​​will be either insufficient or excessive relative to the necessary and sufficient data range. For example, the reactor internal temperature and agitation speed may be selected appropriately, but the raw material composition and catalyst flow rate may be omitted from the selection, resulting in the inclusion of raw material flow rates that need not be considered. It is difficult to improve the prediction accuracy of a prediction model trained on such insufficient and excessive data.

[0023] On the other hand, by using condition information, it is possible to select a data set with a necessary and sufficient range of data, reflecting the type of unit operation, whether the operating method is continuous or batch, data related to start-up and shut-down times, etc., and to train the predictive model. This makes it possible to improve the prediction accuracy of the predictive model. Furthermore, using condition information to select the data to be input into the predictive model contributes to improving the efficiency of analysis and improving the analytical accuracy by using only the necessary data.

[0024] 4 is a block diagram showing the configuration of the plant management device. As shown in FIG. 4, the plant management device includes a computing device 11, a main memory device 12, a user interface 13, a display unit 14, a communication interface 15, and an auxiliary memory device 20.

[0025] The arithmetic device 11 is, for example, a CPU (Central Processing Unit), and operates as a arithmetic unit as defined in the claims. The main storage device 12 is, for example, a RAM (Random Access Memory). The arithmetic unit 11 loads various programs and data into the main storage device 12 and sequentially executes the programs to control the operation of the plant management device.

[0026] The user interface 13 is, for example, a keyboard, and receives operation input from the administrator. Specifically, the user interface 13 is used to input data of the product, etc. The display unit 14 is, for example, a liquid crystal display, and is used to display and output various types of information to the administrator.

[0027] The communication interface 15 communicates with the chemical plant equipment via the data integration platform. In FIG. 4, chemical plants are constructed at multiple locations, such as site L1 and site L2. Various types of plant equipment and sensor devices are installed in the chemical plants at each site. For example, site L1 employs a continuous system and has a set of sensor devices and plant equipment for each unit operation. One or more unit operations form a process, and multiple processes form a production line for the target product.

[0028] The communication interface 15 can receive sensor data from the sensor devices at the base and transmit control instructions to the plant equipment. That is, the communication interface 15 operates as an input unit as defined in the claims with respect to the sensor data.

[0029] The auxiliary storage device 20 is a storage unit that stores various data and operates as a storage unit described in the claims. The auxiliary storage device 20 stores log data 21, a condition information database 22, a unit operation database 23, a prediction model database 24, a process model database 25, reference information 26, anomaly-related information 27, and the like.

[0030] The log data 21 is log data related to the processing of unit operations, for example, operation data of a chemical plant. Each piece of log data may be stored in association with information on the associated unit operation. The condition information database 22 is a database in which condition information relating to the conditions for each unit operation is registered. Condition information may be a mathematical expression, theory, or method of physical or chemical phenomena related to the unit operation being analyzed. Physical or chemical phenomena expressed as mathematical expressions, theories, or methods are stored as mathematical models. Examples include mathematical models that show reaction rate equations, the Arrhenius equation, heat transfer equations, heat transfer area equations, distillation column tray number equations, pressure loss equations, and agitation power equations.

[0031] The unit operation database 23 is a database in which information on a plurality of different unit operations included in the processing steps in a chemical plant is registered. The prediction model database 24 is a database in which prediction models for predicting the results of processing in unit operations are registered. The process model database 25 is a database in which process models for simulating processes including unit operations on a computer are registered. The reference information 26 is reference information indicating that an abnormality has occurred in each unit operation. The abnormality-related information 27 is information in which the operating conditions when an abnormality occurs in a unit operation are stored in association with the unit operation.

[0032] The calculation device 11 creates a prediction model by referring to the sensor data acquired from the sensor device, the log data 21, and the condition information database 22, and registers the prediction model in the prediction model database 24 in association with the unit operation.

[0033] Furthermore, the computing device 11 can create multiple property groups, each containing one or more unit operations having properties within a predetermined range, based on the properties of the unit operations. Then, from among the multiple property groups, the computing device 11 can identify a property group containing a unit operation related to the input sensor data, and create a prediction model based on the sensor data and the log data and condition information related to the one or more unit operations included in the property group.

[0034] For example, information on the type of unit operation may be used as the property of the unit operation. In this case, the arithmetic device 11 creates a plurality of property groups based on the information on the type of unit operation.

[0035] It is also possible to create a property group so that two or more unit operations contained in different types of processes belong to the same group. For example, if the production process for product A and the production process for product B both include "mixing" as a unit operation, create a property group that includes mixing of product A and mixing of product B. Since different processes may share common analysis policies and operating methods, by having two or more unit operations contained in different types of processes belong to the same group, it becomes easier to secure the data necessary for analysis, even in plants that handle a wide variety of products.

[0036] Furthermore, when information on unit operations and condition information at multiple different locations is stored in memory, a property group may be created so that two or more unit operations at different locations belong to the same property group. For example, a property group may be created that includes the mixing when product A is produced at location L1 and the mixing when product A is produced at location L2. Since different locations may have common analysis policies and operating methods, by having two or more unit operations at different locations belong to the same property group, it becomes easier to secure the data necessary for analysis even in plants that handle a wide variety of products.

[0037] Furthermore, characteristic groups may be created based on representative feature quantities for unit operations. This is because when important feature quantities are common, common analysis policies and operating methods are included in the creation of prediction models. For example, characteristic groups may be created so that unit operations for which temperature is an important operating condition belong to the same characteristic group.

[0038] The calculation device 11 can also update the condition information. When performance data, which is the result of processing of a unit operation, is added to the log data 21, the condition information can be updated based on the log data 21 after the addition and the prediction model.

[0039] The computing device 11 can also output operating conditions for unit operations based on the created prediction model and sensor data. For example, the computing device 11 may be configured to output information recommending changes to operating conditions based on information on the results of predictions made by inputting a data set created based on condition information into a prediction model and information on target quality. Optimal conditions may also be searched for based on the prediction model, sensor data, and information on target quality.

[0040] Furthermore, the calculation device 11 can input information on operating conditions into the process model and perform calculations to output information predicting the results of treatment in the unit operation. In this way, by simulating treatment using the process model, various evaluations can be made before actual treatment is performed.

[0041] The computing device 11 can also identify suitable operating conditions for the unit operation based on information predicting the results of processing in the unit operation, create control instructions for plant equipment related to the unit operation, which control instructions include the identified operating conditions, and transmit the control instructions to the plant equipment. By transmitting control instructions in this manner, automatic operation of the plant can be realized.

[0042] Furthermore, the user interface 13 can accept input of any operating conditions for the unit operation, and the arithmetic device 11 can input the received information on the operating conditions into the process model and perform calculations to output the results of processing in the unit operation, and determine whether the output results are abnormal or not based on the reference information 26. If the arithmetic device 11 determines that there is an abnormality, it associates the operating conditions with the unit operation and stores them in the abnormality-related information 27. In this way, by simulating the occurrence of an abnormality, it is possible to accumulate conditions under which an abnormality occurs.

[0043] 5 is an explanatory diagram of the operation of the plant management device. The plant management device performs the following operations in this order: understand unit operation S101, set objective variable S102, select data S103, add feature S104, set learning period S105, and execute soft sensor analysis S106.

[0044] In understanding unit operation S101, the AI ​​learns the conditions registered in the condition information database 22. Specifically, learning by the AI ​​may be realized by calculation by the calculation device 11. The objective variable setting S102 accepts the specification of the analysis target. For example, if the quality of the product is specified as the analysis target, the quality of the product becomes the objective variable. In this case, the prediction model becomes a software sensor that acquires the quality of the product through software calculation.

[0045] In data selection S103, the AI ​​automatically selects data that is highly relevant to the subject of analysis from both a data analysis perspective and condition information perspective. In feature addition S104, the AI ​​automatically adds features that are effective for the soft sensing of the analysis target based on the condition information. In the learning period setting S105, the AI ​​automatically selects an analysis period that is effective for the soft sensing of the analysis target based on the condition information. In the soft sensor analysis execution S106, the AI ​​learns or predicts the target analysis target based on the selected data and period.

[0046] FIG. 6 is an explanatory diagram of a specific example of grouping. When grouping based on unit operations, it is possible to group reactions such as "grouping reactions with similar theoretical models together" or "integrating fractional distillation, general distillation, and reduced pressure distillation into groups based on distillation classification." When grouping based on characteristics, it is possible to "group by fluid (solid, liquid, gas)." For example, "unit operation α, liquid" can be grouped into Gr11, and "unit operation α, gas" can be grouped into Gr12. It is also possible to "group by the equipment used." For example, it is possible to group by whether the reaction is performed continuously or batchwise.

[0047] Next, the operation of the plant management device will be described. FIG. 7 is a flowchart showing the operation when creating a prediction model. The arithmetic device 11 first executes a sensor data input step of inputting sensor data acquired from a sensor device in the plant (step S201).

[0048] Thereafter, the arithmetic device 11 uses the sensor data, the log data 21, and the condition information in the condition information database 22 to generate a data set to be used for creating a prediction model (step S202).

[0049] The computing device 11 performs machine learning using the generated data set to create a prediction model for predicting the results of processing in the unit operation (step S203). With the above configuration, it is possible to create a model for analyzing the sensor information of the analysis target while taking into account constraints such as the analysis policy and analysis conditions specific to the unit operation. The created prediction model is then stored in the prediction model database 24 of the auxiliary storage device 20 in association with the unit operation (step S204), and the process ends. If the amount of accumulated sensor data for a particular unit operation is sufficient, a prediction model can be created by inputting only the sensor data for that unit operation in step S201. On the other hand, even if the amount of accumulated sensor data for that particular unit operation is insufficient, a prediction model can be created by using data accumulated for other unit operations in the group to which that unit operation belongs.

[0050] FIG. 8 is a flowchart showing the operation of the prediction model when it is in operation. The computing device 11 first executes a sensor data input step of inputting sensor data acquired from a sensor device in the plant (step S301). Here, when the sensor information to be analyzed is related to an arbitrary first unit operation, the input sensor data (sensor information) may be determined based on condition information related to the first unit operation. This allows data for creating a dataset required for analysis to be acquired efficiently and reliably.

[0051] Thereafter, the computing device 11 generates a data set to be given to the prediction model using the sensor data and the condition information in the condition information database 22 (step S302). Since the type, period, etc. of sensor data required for analysis differs for each unit operation, generating a data set to be given to the prediction model based on the condition information makes it possible to use only the necessary data and perform more accurate and efficient predictions.

[0052] The computing device 11 executes prediction using a prediction model with the generated data set as explanatory variables (step S303). At this time, the prediction model used may be a prediction model created only from the data of the target unit operation, or a prediction model created using data of the group to which the target unit operation belongs. Then, the prediction result is output (step S304), and the process ends.

[0053] Figure 9 is an explanatory diagram of automatic plant operation. The plant management device generates a data set using the sensor data and condition information acquired by the sensor device. By inputting this data set into a prediction model to simulate the state of the reactor, it is possible to predict the actual state of the reactor. Based on the prediction results, the plant management device identifies optimal operating conditions for the unit operations, generates control instructions for the plant, and sends them to the plant equipment. By sending control instructions in this manner, automatic plant operation can be achieved.

[0054] Although Figure 4 shows a state in which the plant at site L1 employs a continuous system, the plant may also employ a batch system. Figure 10 shows a case in which site L1 employs a batch system. Because a batch system is employed, multiple unit operations are performed in one plant facility at site L1, and then the product is transferred to the facility for the next process. The sensor devices are installed in association with the plant facility. In other words, one sensor device may be used in multiple unit operations. For this reason, the output of the sensor device must be accompanied by information that can identify which unit operation the sensor output is for, and transmitted to the plant management device. The other configurations and operations are the same as those in Figure 4, and therefore will not be described here.

[0055] Although the above description has exemplified a method of dividing unit operations into multiple groups to create a prediction model, other methods can also be used. For example, the computing device 11 can identify a second unit operation that is different from the first unit operation and related to the first unit operation, and then create a prediction model for predicting the results of processing in the first unit operation based on log data information related to the second unit operation and condition information related to the second unit operation. As an example, by searching for unit operations similar to a certain unit operation and creating a prediction model using accumulated log data and condition information about the similar unit operations, the amount of data available for creating the prediction model can be increased.

[0056] As described above, the disclosed plant management device comprises an auxiliary storage device 20 as a storage unit that stores information on multiple different unit operations included in a processing process in a chemical plant, log data information related to the processing of the unit operations, and condition information related to the conditions for each of the unit operations; a communication interface 15 as an input unit that inputs sensor information obtained from a sensor device related to a first unit operation in the chemical plant; and a calculation device 11 as a calculation unit that creates a prediction model for predicting the results of processing in the first unit operation based on the sensor information, the log data information related to the first unit operation, and the condition information, and stores the prediction model in the storage unit in association with the first unit operation. By adopting such a configuration and operation, the plant management device can predict the results of processing by a chemical plant, reflecting the characteristics of the unit operations.

[0057] The calculation unit can also create a plurality of characteristic groups, each containing one or more unit operations having characteristics within a predetermined range, based on the characteristics of the unit operations, identify a first characteristic group from among the plurality of characteristic groups that contains the first unit operation associated with the sensor information input to the input unit, and create a prediction model for predicting the results of processing in the first unit operation based on the sensor information, the information in the log data associated with one or more unit operations included in the first characteristic group, and the condition information. Therefore, even in plants that handle a wide variety of products in small quantities, it is possible to create a prediction model efficiently.

[0058] The condition information may also include information on a predetermined mathematical model associated with each of the unit operations. Furthermore, the characteristics of the unit operation may be information on the type of the unit operation, and the calculation unit may create the plurality of characteristic groups based on the information on the type of the unit operation. Furthermore, the property group may include two or more unit operations that are respectively included in different types of processes in the same group. In addition, the memory unit stores information on the unit operations and the condition information at multiple different locations, and the characteristic group may include two or more unit operations at different locations in the same group. In addition, the characteristics of the unit operation may be information indicating a representative feature of the unit operation, and the calculation unit may create the multiple characteristic groups based on the information indicating the representative feature. In this way, any condition can be used as the condition information, and the unit operations can be grouped in any way.

[0059] In addition, the memory unit may further store first performance data which is the result of processing the first unit operation, and the calculation unit may update the condition information regarding the first unit operation based on the created prediction model and the first performance data. In this way, by updating the condition information using data accumulated through operation, the accuracy of predictions made by the prediction model can be improved.

[0060] The calculation unit may also output first operating conditions for the first unit operation based on the created prediction model and the sensor information. Furthermore, the memory unit may store a process model that simulates a process including the first unit operation on a computer, and the calculation unit may input information on the first operating conditions into the process model and perform calculations to output information that predicts the results of processing in the first unit operation. Furthermore, the calculation unit may identify second operating conditions for the first unit operation based on information predicting the processing results for the first unit operation, create control instructions for plant equipment related to the first unit operation, which control instructions include the second operating conditions, and transmit the control instructions to the plant equipment. In this way, predictions made by the predictive model can be used to output appropriate operating conditions, simulate processes, and perform automated operations.

[0061] The memory unit may also store reference information, which is standard information indicating that an abnormality has occurred for each of the unit operations, and abnormality-related information, which is information storing the operating conditions when an abnormality occurred in the unit operation in association with the unit operation; the input unit may be capable of accepting input of any second operating conditions for the first unit operation; the calculation unit may input information on the second operating conditions into the process model and perform calculations to output the results of processing in the first unit operation, determine whether the output result is abnormal based on the reference information, and if it is determined to be abnormal, store the second operating conditions in association with the first unit operation in the abnormality-related information. By adopting such a configuration and operation, it is possible to simulate abnormal states and accumulate conditions under which abnormalities occur.

[0062] The calculation unit may also identify a second unit operation that is different from the first unit operation and related to the first unit operation, and create a prediction model for predicting the results of processing in the first unit operation based on information in the log data related to the second unit operation and the condition information related to the second unit operation. In this way, similar unit operations can be identified by any method, without being limited to grouping, and a prediction model can be created from information related to different unit operations.

[0063] From another perspective, the disclosed plant management device comprises an auxiliary storage device 20 as a storage unit that stores information on a plurality of unit operations included in a processing step in a chemical plant, information on log data related to the processing of the unit operations, and condition information related to the conditions for each of the unit operations; an arithmetic unit 11 as a processing unit that creates a plurality of property groups each including one or more unit operations having properties within a predetermined range based on the properties of the unit operations; and a communication interface 15 as an input unit that inputs sensor information obtained from a sensor device related to a first unit operation in the chemical plant, and the arithmetic unit 11 also operates as a calculation unit that identifies a first property group from the plurality of property groups that includes the first unit operation related to the sensor information input to the input unit, and predicts the results of the processing in the first unit operation based on the sensor information, the log data information related to one or more unit operations included in the first property group, and the condition information. Therefore, the plant management device can predict the results of processing in a chemical plant, reflecting the characteristics of the unit operations.

[0064] The disclosed plant management method also includes the steps of inputting sensor information acquired from a sensor device related to a first unit operation included in a plurality of different unit operations included in a processing process in a chemical plant, creating a predictive model for predicting the results of processing in the first unit operation based on the sensor information, log data information related to the processing of the first unit operation, and condition information which is information related to the conditions in the first unit operation, and storing the predictive model in a specified storage device in association with the first unit operation. Therefore, the plant management method can predict the results of processing by a chemical plant, reflecting the characteristics of the unit operations.

[0065] The present invention is not limited to the above-described embodiment, but includes various modifications. For example, the above-described embodiment has been described in detail to clearly explain the present invention, and is not necessarily limited to an embodiment including all of the described configurations. Furthermore, configurations can be replaced or added, not limited to deletion. [Explanation of symbols]

[0066] 11: arithmetic unit, 12: main memory device, 13: user interface, 14: display unit, 15: communication interface, 20: auxiliary memory device, 21: log data, 22: condition information database, 23: unit operation database, 24: prediction model database, 25: process model database, 26: reference information, 27: abnormality-related information

Claims

1. A storage unit that stores information on a plurality of different unit operations included in a processing step in a chemical plant, information on log data related to the processing of the unit operations, and condition information related to the conditions for each of the unit operations; an input unit for inputting sensor information acquired from a sensor device related to a first unit operation in a chemical plant; a calculation unit that creates a prediction model for predicting the processing result of the first unit operation based on the sensor information, the log data information related to the first unit operation, and the condition information, and stores the prediction model in the storage unit in association with the first unit operation; The calculation unit creating a plurality of characteristic groups, each containing one or more of the unit operations, based on the characteristics of the unit operations; a first characteristic group including the first unit operation related to the sensor information input to the input unit is identified from among the plurality of characteristic groups, and a prediction model is created for predicting the processing result of the first unit operation based on the sensor information, the log data information related to one or more unit operations included in the first characteristic group, and the condition information; The characteristics of the unit operation are information on the type of the unit operation, The calculation unit creates the plurality of characteristic groups based on information on the type of the unit operation, Furthermore, the characteristic group includes two or more unit operations each included in a different type of process in the same group.

2. The plant management device according to claim 1, A plant management device in which the condition information includes information on a predetermined mathematical model associated with each of the unit operations.

3. The plant management device according to claim 1, the storage unit further stores first performance data which is a result of processing the first unit operation, The calculation unit updates the condition information regarding the first unit operation based on the created prediction model and the first performance data.

4. The plant management device according to claim 1, the calculation unit outputs first operating conditions for a first unit operation based on the created prediction model and the sensor information, the storage unit stores a process model that simulates a process including the first unit operation on a computer, the calculation unit inputs information on the first operating condition into the process model and performs calculations to output information predicting a result of the treatment in the first unit operation, the calculation unit identifies second operating conditions for the first unit operation based on information predicting the result of the treatment in the first unit operation; A plant management device that creates control instructions for plant equipment related to the first unit operation, the control instructions including the second operating conditions, and transmits the control instructions to the plant equipment.

5. The plant management device according to claim 1, The sensor information acquired from the sensor device related to the first unit operation and input to the input unit is determined based on the condition information related to the first unit operation. This is a plant management device.

6. A memory unit that stores information on a plurality of different unit operations included in a processing process in a chemical plant, information on log data related to the processing of the unit operations, and condition information related to the conditions for each of the unit operations; an input unit for inputting sensor information acquired from a sensor device related to a first unit operation in a chemical plant; a calculation unit that creates a prediction model for predicting the processing result of the first unit operation based on the sensor information, the log data information related to the first unit operation, and the condition information, and stores the prediction model in the storage unit in association with the first unit operation; The calculation unit creating a plurality of characteristic groups, each containing one or more of the unit operations, based on the characteristics of the unit operations; a first characteristic group including the first unit operation related to the sensor information input to the input unit is identified from among the plurality of characteristic groups, and a prediction model is created for predicting the processing result of the first unit operation based on the sensor information, the log data information related to one or more unit operations included in the first characteristic group, and the condition information; The storage unit stores information on the unit operations and the condition information at a plurality of different base stations, Furthermore, the plant management device is one in which the characteristic group includes two or more unit operations at different bases in the same group.

7. A memory unit that stores information on a plurality of different unit operations included in a processing process in a chemical plant, information on log data related to the processing of the unit operations, and condition information related to the conditions for each of the unit operations; an input unit for inputting sensor information acquired from a sensor device related to a first unit operation in a chemical plant; a calculation unit that creates a prediction model for predicting the processing result of the first unit operation based on the sensor information, the log data information related to the first unit operation, and the condition information, and stores the prediction model in the storage unit in association with the first unit operation; The calculation unit creating a plurality of characteristic groups, each containing one or more of the unit operations, based on the characteristics of the unit operations; a first characteristic group including the first unit operation related to the sensor information input to the input unit is identified from among the plurality of characteristic groups, and a prediction model is created for predicting the processing result of the first unit operation based on the sensor information, the log data information related to one or more unit operations included in the first characteristic group, and the condition information; The characteristics of the unit operation are information indicating a representative feature amount in the unit operation, The calculation unit creates the plurality of characteristic groups based on information indicating the representative feature amounts.

8. A memory unit that stores information on a plurality of different unit operations included in a processing process in a chemical plant, information on log data related to the processing of the unit operations, and condition information related to the conditions for each of the unit operations; an input unit for inputting sensor information acquired from a sensor device related to a first unit operation in a chemical plant; a calculation unit that creates a prediction model for predicting the processing result of the first unit operation based on the sensor information, the log data information related to the first unit operation, and the condition information, and stores the prediction model in the storage unit in association with the first unit operation; the calculation unit outputs first operating conditions for a first unit operation based on the created prediction model and the sensor information, the storage unit stores a process model that simulates a process including the first unit operation on a computer, the calculation unit inputs information on the first operating condition into the process model and performs calculations to output information predicting a result of the treatment in the first unit operation, The storage unit stores reference information, which is reference information indicating that an abnormality has occurred in each of the unit operations, and abnormality-related information, which is information storing the operating conditions when an abnormality has occurred in the unit operation in association with the unit operation; The input unit is capable of accepting input of any second operating condition for the first unit operation, The calculation unit inputs information on the second operating conditions into the process model and performs calculations to output the results of processing in the first unit operation, determines whether the output result is abnormal based on the reference information, and if it is determined to be abnormal, associates the second operating conditions with the first unit operation and stores them in the abnormality-related information. This is a plant management device.

9. A memory unit that stores information on a plurality of different unit operations included in a processing process in a chemical plant, information on log data related to the processing of the unit operations, and condition information related to the conditions for each of the unit operations; an input unit for inputting sensor information acquired from a sensor device related to a first unit operation in a chemical plant; a calculation unit that creates a prediction model for predicting the processing result of the first unit operation based on the sensor information, the log data information related to the first unit operation, and the condition information, and stores the prediction model in the storage unit in association with the first unit operation; The calculation unit identifying a second unit operation that is different from the first unit operation and that is associated with the first unit operation; The plant management device further creates a predictive model for predicting the results of processing in the first unit operation based on the log data information related to the second unit operation and the condition information related to the second unit operation.

10. A storage unit that stores information on a plurality of unit operations included in a processing step in a chemical plant, information on log data related to the processing of the unit operations, and condition information related to the conditions for each of the unit operations; a processing unit that creates a plurality of characteristic groups based on the characteristics of the unit operations, each group including one or more unit operations having a predetermined range of characteristics; an input unit for inputting sensor information acquired from a sensor device related to a first unit operation in a chemical plant; a calculation unit that identifies a first property group among the plurality of property groups, which includes the first unit operation related to the sensor information input to the input unit, and predicts a processing result of the first unit operation based on the sensor information, the log data information related to one or more unit operations included in the first property group, and the condition information; The characteristics of the unit operation are information on the type of the unit operation, The calculation unit creates the plurality of characteristic groups based on information on the type of the unit operation, Furthermore, the characteristic group includes two or more unit operations each included in a different type of process in the same group.

11. inputting sensor information acquired from a sensor device associated with a first unit operation included in a plurality of different unit operations included in a process in a chemical plant; A step of creating a prediction model for predicting the results of the processing in the first unit operation based on the sensor information, information on log data related to the processing of the first unit operation, and condition information which is information on the conditions in the first unit operation; and storing the prediction model in a predetermined storage device in association with the first unit operation; further comprising creating a plurality of characteristic groups based on the characteristics of the unit operations, each group including one or more of the unit operations; The step of creating the prediction model involves identifying a first characteristic group from the plurality of characteristic groups that includes the first unit operation related to the sensor information, and creating a prediction model for predicting the processing result of the first unit operation based on the sensor information and log data information and condition information related to one or more of the unit operations included in the first characteristic group; The characteristics of the unit operation are information on the type of the unit operation, The step of creating the plurality of characteristic groups creates the plurality of characteristic groups based on information on the type of the unit operation, Furthermore, in the plant management method, the characteristic group includes two or more unit operations each included in a different type of process in the same group.

Images