Information display device, information display method, and program

Abstract

To provide technology that can present procedures for responding to specific events. [Solution] An information presentation device according to one aspect of the present disclosure includes: a construction unit that constructs a database containing data that includes text information representing the content of a document and storage location information indicating the location where the document is stored, based on a document relating to the operation of a target; a generation unit that, when a specific event occurs in the target, generates a response procedure for the event based on text information contained in the data corresponding to the event among the data constituting the database; and a presentation unit that presents to the user operating the target information that the event occurred, the response procedure, and the document stored in the storage location indicated by the storage location information contained in the data.

Description

【Technical Field】 【0004】 , 【0006】 , , , , , , , , , 【0005】 【0001】 The present disclosure relates to an information presentation device, an information presentation method, and a program. 【Background Art】 【0002】 When an abnormality or its sign occurs in a plant, equipment, etc., an operation to recover from the abnormality or avoid the abnormality is required. For this reason, a technique for presenting operation procedures for a specific event to an operator when the specific event indicating the abnormality or its sign occurs in a plant, equipment, etc. is known (for example, Patent Documents 1 to 3). 【Prior Art Documents】 【Patent Documents】 【0003】 【Patent Document 1】 Japanese Unexamined Patent Application Publication No. 2023-111624 【Patent Document 2】 Japanese Unexamined Patent Application Publication No. 2021-64199 【Patent Document 3】 Japanese Unexamined Patent Application Publication No. 2020-21108 【Summary of the Invention】 【Problems to be Solved by the Invention】 【0004】 However, in the prior art, it requires an operator's input (e.g., input of information representing an event that has occurred in a plant, equipment, etc.) or past operation performance data of a plant, equipment, etc. For this reason, a technique that can present operation procedures for a specific event without requiring an operator's input and without necessarily requiring past operation performance data is demanded. 【0005】 The present disclosure has been made in view of the above points, and an object thereof is to provide a technique that can present operation procedures for a specific event. 【Means for Solving the Problems】 【0006】 An information presentation device according to one aspect of the present disclosure includes: a construction unit that constructs a database containing data that includes text information representing the content of a document and storage location information indicating the location where the document is stored, based on a document relating to the operation of the target; a generation unit that, when a specific event occurs in the target, generates a response procedure for the event based on text information contained in the data corresponding to the event among the data constituting the database; and a presentation unit that presents to the user operating the target information that the event occurred, the response procedure, and the document stored in the storage location indicated by the storage location information contained in the data. [Effects of the Invention] 【0007】 A technology is provided that can present procedures for responding to specific events. [Brief explanation of the drawing] 【0008】 [Figure 1] This figure shows an example of the overall configuration of the driver assistance system according to this embodiment. [Figure 2] This figure shows an example of measurement data stored in the measurement data database. [Figure 3] This figure shows an example of operation log data stored in the operation log database. [Figure 4] This flowchart shows an example of the operation procedure for building a database according to this embodiment. [Figure 5] This figure shows an example of text data obtained from technical documents. [Figure 6] This figure shows an example of a list of corresponding operations. [Figure 7] This figure shows an example of supplementary information. [Figure 8] This diagram shows an example of the corrective operation procedure. [Figure 9] This flowchart shows an example of the abnormality monitoring process according to this embodiment. [Figure 10] This figure shows an example of a prompt template. [Figure 11] This figure shows an example of a prompt. [Figure 12] It is a diagram (part 1) showing an example of a prompt screen. [Figure 13] It is a flowchart showing an example of the inquiry response process according to the present embodiment. [Figure 14] It is a diagram (part 2) showing an example of a prompt screen. [Figure 15] It is a diagram showing an example of the hardware configuration of a computer. 【Mode for Carrying Out the Invention】 【0009】 Hereinafter, an embodiment of the present invention will be described in detail with reference to the drawings. In the following embodiments, a driving support system 1 will be described that can present a driver with response procedures for specific events that occurred in a plant without requiring the input of a driver and without necessarily requiring past driving performance data. Here, examples of specific events include plant abnormalities or their omens. Note that the plant is not limited to a specific plant, and various arbitrary plants (e.g., waste incineration plants, steel plants, petrochemical plants, power generation plants, food plants, etc.) can be targeted. 【0010】 However, targeting a plant is just an example, and it is also possible to target various other facilities, devices, apparatuses, etc. other than plants. 【0011】 <Overall configuration example of the driving support system 1> FIG. 1 is a diagram showing an example of the overall configuration of the driving support system 1 according to the present embodiment. As shown in FIG. 1, the driving support system 1 according to the present embodiment includes a support information presentation system 10, a driver device 20, and a plant system 30. Here, the support information presentation system 10 and the driver device 20 are communicably connected via an arbitrary communication network. Similarly, the driver device 20 and the plant system 30 are communicably connected via an arbitrary communication network. 【0012】 When an abnormality or its sign occurs in the plant, the support information presentation system 10 transmits an abnormality occurrence notice indicating the occurrence to the operator device 20, or transmits a corresponding procedure for the abnormality or its sign and reference materials regarding the corresponding procedure to the operator device 20. Further, when the support information presentation system 10 receives a certain inquiry (e.g., an inquiry regarding the corresponding procedure, etc.) from the operator device 20, it generates an answer to the inquiry and then transmits the answer and reference materials regarding the answer. 【0013】 The operator device 20 is a terminal device used by the operator of the plant. The operator device 20 receives an abnormality detection notice, a corresponding procedure, and reference materials from the support information presentation system 10. Further, the operator device 20 transmits an inquiry to the support information presentation system 10 or receives an answer to the inquiry and reference materials from the support information presentation system 10. 【0014】 The plant system 30 is a system that realizes the plant. Here, the plant system 30 includes one or more plant facilities 301, one or more sensors 302, a data collection device 303, a measurement value DB 304, and an operation log DB 305. 【0015】 Plant equipment 301 consists of various pieces of equipment that perform the plant's processes (e.g., waste incineration in the case of a waste incineration plant, iron and steel production in the case of a steel plant). Sensor 302 is a device that measures values ​​(hereinafter also referred to as "measured values") of certain objects in the plant equipment 301 or processes (e.g., furnace gas temperature, furnace gas pressure, furnace gas flow rate, feeder speed, orifice damper opening, ON / OFF of some setting, etc.). Data acquisition device 303 collects the measured values ​​measured by sensor 302 and logs (hereinafter referred to as "operation logs") related to various operations on the plant equipment 301, alarm occurrences, or recovery. Data acquisition device 303 also stores the measured value data, which includes each measured value, in the measured value DB 304, and the operation log data, which includes the operation logs, in the operation log DB 305. Measured value DB 304 is a database composed of measured value data. Operation log DB 305 is a database composed of operation log data. Hereafter, the object whose value is measured by sensor 302 will be referred to as the "measurement target". 【0016】 <<Example Configuration of Support Information Presentation System 10>> The support information presentation system 10 according to this embodiment includes an operation procedure database construction device 110, an abnormality monitoring device 120, and an operation guidance device 130. Here, the operation procedure database construction device 110 and the operation guidance device 130 are connected to communicate via an arbitrary communication network. Similarly, the abnormality monitoring device 120 and the operation guidance device 130 are connected to communicate via an arbitrary communication network. 【0017】 However, for example, the operation procedure database construction device 110, the anomaly monitoring device 120, and the operation guidance device 130 may be integrated to constitute the support information presentation system 10. 【0018】 • Example configuration of the operation procedure DB construction device 110 The operation procedure DB construction device 110 includes a text acquisition unit 111, a text conversion unit 112, and a log analysis unit 113. Each of these units is implemented, for example, by a process executed by a computing device such as a CPU (Central Processing Unit) through one or more programs installed in the operation procedure DB construction device 110. The operation procedure DB construction device 110 also includes an operation procedure DB 114 and a log analysis result DB 115. Each of these DBs is implemented, for example, by the storage area of ​​a storage device such as an HDD (Hard Disk Drive), SSD (Solid State Drive), or flash memory. 【0019】 The text acquisition unit 111 takes the given technical document 400 as input and acquires text data representing the text in the technical document 400. The technical document 400 is a document that is used as a reference when operating the plant. Specific examples of the technical document 400 include, for example, an operation manual and training materials. Other specific examples of the technical document 400 include, for example, a collection of case studies describing abnormalities or signs of abnormalities that have occurred in the plant in the past and the corresponding response procedures, and an inspection manual for when abnormalities or signs of abnormalities occur. 【0020】 The text conversion unit 112 converts the text data acquired by the text acquisition unit 111 into vector data. The text conversion unit 112 also creates operation procedure data that associates the vector data with the original text data from which the vector data was converted and the storage location of the technical document 400 from which the text data was acquired, and then stores the operation procedure data in the operation procedure DB 114. 【0021】 Similarly, the text conversion unit 112 converts text data representing text in the log analysis results from the log analysis unit 113 into vector data. The text conversion unit 112 also creates operation procedure data that associates the vector data with the source text data from which the vector data was converted and the storage location of the log analysis results corresponding to the text data, and then stores the operation procedure data in the operation procedure DB 114. Hereinafter, the data representing the log analysis results will also be referred to as "log analysis result data". 【0022】 The log analysis unit 113 analyzes operation log data stored in the operation log DB 305 if such data exists. The log analysis unit 113 also stores the log analysis result data, which is the result of the operation log data analysis, in the log analysis result DB 115. 【0023】 Operation Procedure DB114 is a database composed of operation procedure data. Operation procedure data is data that associates vector data with the text data from which that vector data was converted, and with the storage location of the technical document 400 or log analysis result data that contains the text represented by that text data. 【0024】 The log analysis results DB115 is a database composed of log analysis result data. Log analysis result data is data that represents the results of analyzing the operations, their sequence, number of times, etc., that occurred when an anomaly or its indication occurred, using operation log data from the time the anomaly or its indication occurred until the response to the anomaly or its indication was completed. For example, log analysis result data may include a list of corresponding operations that shows the operations performed and their number of times from the time the anomaly or its indication occurred until the anomaly or its indication was resolved. Other examples of log analysis result data may include a list of corresponding operations that show the operations performed and their sequence from the time the anomaly or its indication occurred until the anomaly or its indication was resolved. 【0025】 • Example configuration of the abnormality monitoring device 120 The anomaly monitoring device 120 includes an anomaly detection unit 121 and a prompt creation unit 122. Each of these units is implemented, for example, by a process in which one or more programs installed in the anomaly monitoring device 120 are executed by a processing unit such as a CPU. 【0026】 The anomaly detection unit 121 detects the occurrence of an anomaly in the plant or a sign thereof based on at least one of the measurement value data stored in the measurement value DB 304 and the operation log data stored in the operation log DB 305. 【0027】 The prompt generation unit 122 generates a prompt based on the operation log data at the time the anomaly detection unit 121 detects the occurrence of an anomaly or its signs. A prompt is information (especially information written in natural language) given as input to a trained machine learning model, such as a Large Language Model (LLM). A prompt may also be called, for example, a "question," "instruction," or "command." A Large Language Model can be realized, for example, by training a machine learning model that includes a neural network such as a Transformer using a large amount of text data. 【0028】 • Example configuration of the operation guidance device 130 The operation guidance device 130 includes a text search unit 131 and a text generation unit 132. Each of these units is realized, for example, by having one or more programs installed in the operation guidance device 130 execute processes on a processing unit such as a CPU. 【0029】 The text search unit 131 converts the prompt created by the prompt creation unit 122 into vector data, and then searches for operation procedure data stored in the operation procedure DB 114 based on this vector data. 【0030】 Similarly, when the text search unit 131 receives a query from the operator device 20, it acquires the query as a prompt and converts the prompt into vector data. The text search unit 131 then searches for operation procedure data stored in the operation procedure DB 114 based on this vector data. 【0031】 The text generation unit 132 generates a corresponding procedure using a large-scale language model based on the prompt created by the prompt creation unit 122 and the text data contained in the operation procedure data retrieved by the text search unit 131. The text generation unit 132 also sends an abnormality notification indicating that an abnormality or its signs have occurred, along with a corresponding procedure for that abnormality or its signs, to the operator device 20, and also sends data stored in the storage location included in the operation procedure data (technical document 400 or log analysis result data) to the operator device 20 as reference material. 【0032】 Similarly, the text generation unit 132 generates an answer to the query using a large-scale language model based on the prompt received from the operator device 20 and the text data contained in the operation procedure data retrieved by the text search unit 131. The text generation unit 132 also transmits the answer to the operator device 20 and transmits the data stored in the storage location included in the operation procedure data (technical document 400 or log analysis result data) to the operator device 20 as reference material. 【0033】 ≪Example of measurement data stored in Measurement Value DB304≫ Figure 2 shows an example of measurement data stored in the measurement data DB 304. As shown in Figure 2, the measurement data DB 304 stores one or more measurement data, and each measurement data includes "date and time" and "tag" as data items. The data item "date and time" is set to the date and time when the measurement value of the object to be measured was measured by the sensor 302 (or the date and time when these measurement values ​​were collected by the data acquisition device 303). The data item "tag" is set to a tag. Here, a tag is information that represents the object to be measured. 【0034】 In the example shown in Figure 2, the measurement period for the measured values ​​is 1 second, and the measured value data is stored in the measured value DB304 for each measurement period. 【0035】 ≪Example of operation log data stored in Operation Log DB305≫ Figure 3 shows an example of operation log data stored in the operation log DB 305. As shown in Figure 3, the operation log DB 305 stores one or more operation log data, and each operation log data includes the following data items: "Date and Time", "Tag", "Event", "Status", and "Alarm". The data item "Date and Time" is set to the date and time when the operation was performed or when an alarm occurred or when recovery from that alarm occurred (or it may be the date and time when the operation log was collected by the data acquisition device 303). The data item "Tag" is set to a tag. The "Event" is set to information that represents the content of the operation log. The "Status" is set to information that represents the content of the operation if the operation log indicates that an operation was performed. The "Alarm" is set to information that represents the occurrence or recovery from an alarm if the operation log indicates the occurrence or recovery from that alarm. 【0036】 In the example shown in Figure 3, the data item "Alarm" in the first row of the operation log data is set to "Occurred," indicating that an alarm occurred (i.e., an anomaly occurred). The data item "Alarm" in the fifth row of the operation log data is set to "Recovered," indicating that the system recovered from the alarm (i.e., recovered from the anomaly). The operation log data in rows 2 through 4 represents the operations performed between the occurrence of the alarm and the recovery (i.e., operations performed to recover from the anomaly). 【0037】 <Example of DB construction process in operation procedure> Figure 4 is a flowchart showing an example of the operation procedure DB construction process according to this embodiment. 【0038】 The text acquisition unit 111 of the operation procedure DB construction device 110 takes the given technical document 400 as input and acquires text data representing the text in the technical document 400 (step S101). For example, if the technical document 400 is data that has been digitized from a paper document using a scanner or the like (e.g., data in the form of a PDF file, image file, etc.), the text acquisition unit 111 can acquire the text data using optical character recognition (OCR). 【0039】 As an example, Figure 5 shows text data obtained from a certain technical document 400. The text data 410 shown in Figure 5 is text data obtained from technical document 400, which describes the response procedure when an abnormality occurs in the lower limit of the furnace gas temperature. The text data 410 shown in Figure 5 states that the causes of the abnormality in the lower limit of the furnace gas temperature are "thin waste layer" and "decreased O2 concentration." The text data 410 shown in Figure 5 also describes the operations of "increasing the feeder speed" and "increasing the opening of the orifice damper" as countermeasures against the abnormality in the lower limit of the furnace gas temperature. 【0040】 The text conversion unit 112 of the operation procedure DB construction device 110 creates vector data by vectorizing the text data obtained in step S101 (step S102). That is, the text conversion unit 112 converts the text data obtained in step S101 into vector data. The text conversion unit 112 can convert the text data into vector data using existing methods that convert unstructured data into fixed-dimensional vector data (e.g., conversion processing using an embedding model). It is assumed that the number of dimensions of the vectors represented by the vector data is predetermined. 【0041】 The text conversion unit 112 of the operation procedure DB construction device 110 creates operation procedure data by associating the vector data created in step S102 with the text data acquired in step S101 and the storage location of the given technical document 400, and then stores the said operation procedure data in the operation procedure DB 114 (step S103). 【0042】 Steps S101 to S103 described above are repeated for each given technical document 400. 【0043】 The log analysis unit 113 of the operation procedure DB construction device 110 determines whether or not operation log data exists in the operation log DB 305 (step S104). 【0044】 If, in step S104 above, it is determined that no operation log data exists stored in the operation log DB 305, the operation procedure DB construction device 110 terminates the operation procedure DB construction process. 【0045】 On the other hand, if it is determined in step S104 above that operation log data exists stored in the operation log DB 305, the log analysis unit 113 of the operation procedure DB construction device 110 retrieves the operation log data stored in the operation log DB 305 (step S105). 【0046】 The log analysis unit 113 of the operation procedure DB construction device 110 uses the operation log data acquired in step S105 to create log analysis result data that represents the results of analyzing the operations, their sequence, the number of times, etc., when an abnormality or its indication occurred (step S106). The log analysis unit 113 can create log analysis result data using methods such as statistical analysis or process mining. 【0047】 ·Statistical analysis As an example, let's describe how log analysis result data is created using statistical analysis. In this case, the log analysis unit 113 can create log analysis result data by following the steps 1-1 to 1-5 below. 【0048】 Procedure 1-1: The log analysis unit 113 extracts operation log data from the operation log data acquired in step S105 above, from the period after the occurrence of an abnormality or its indication until the abnormality or its indication ends. For example, the log analysis unit 113 extracts operation log data from the operation log data where the data item "Alarm" is set to "Occurred" to the operation log data where the data item "Alarm" is set to "Recovered". 【0049】 Step 1-2: The log analysis unit 113 uses the operation log data extracted in Step 1-1 above to calculate the names of the operations performed and the number of times they were performed between the time the abnormality or its indication occurred and the time the abnormality or its indication ended. This creates a corresponding operation list that shows the names of the operations performed and the number of times they were performed between the time the abnormality or its indication occurred and the time the abnormality or its indication ended. 【0050】 The log analysis unit 113 can, for example, use the information set in the data item "Event" of the operation log data, where the data item "Status" contains information representing the content of the operation, as the operation name. 【0051】 Step 1-3: The log analysis unit 113 assigns a data name containing the name of the abnormality or its symptoms to the list of corresponding operations created in Step 1-2 above. For example, the log analysis unit 113 can create the data name by setting the name of the abnormality or its symptoms to the variable part {} of the template "{}List of corresponding operations when an abnormality occurs" (where {} is the variable part). As a result, a list of corresponding operations with data names containing the name of the abnormality or its symptoms is obtained as log analysis result data. 【0052】 As an example, Figure 6 shows the list of corresponding operations when the list of corresponding operations is created as a JPEG image file in step 1-2 above, and the file name "List of corresponding operations when an abnormality occurs at the lower limit of the furnace gas temperature" is created as the data name in step 1-3 above. In the list of corresponding operations 1100 shown in Figure 6, operations A to D were performed from the time an abnormality occurred at the lower limit of the furnace gas temperature until the abnormality was recovered. In addition, the number of times each operation A to D was performed from the time an abnormality occurred at the lower limit of the furnace gas temperature until the abnormality was recovered is shown in a graph in the list of corresponding operations 1100 shown in Figure 6. 【0053】 Step 1-4: The log analysis unit 113 creates supplementary information that includes the collection period of the operation log data obtained in step S105 above, the name of the anomaly or its indication extracted in step 1-1 above, the number of times the anomaly or its indication occurred during the collection period, the average recovery time, and the frequency of each operation when recovery was made from the anomaly or its indication. 【0054】 Step 1-5: The log analysis unit 113 assigns a data name to the supplementary information created in Step 1-4 above, which includes the name of the anomaly or its symptom. For example, the log analysis unit 113 can create the data name by setting the name of the anomaly or its symptom to the variable part {} of the pre-created template "{} Supplementary information when an anomaly occurs" (where {} is the variable part). As a result, the supplementary information with a data name containing the name of the anomaly or its symptom is obtained as log analysis result data. 【0055】 As an example, Figure 7 shows the supplemental information when supplemental information is created as a text file in txt format in step 1-4 above, and the file name is set as "Supplemental Information for Abnormal Lower Limit Occurrence of Furnace Gas Temperature" in step 1-5 above. In the supplemental information 1200 shown in Figure 7, the collection period of the operation log data obtained in step S105 above is listed as "Operation Log Collection Period". Also, in the supplemental information 1200 shown in Figure 7, the name of the abnormality or its symptom extracted in step 1-1 above, the number of times the abnormality or its symptom occurred during the collection period, and the average recovery time from the abnormality or its symptom are listed as "Abnormal Event". Furthermore, in the supplemental information 1200 shown in Figure 7, the operation frequency of each operation when recovering from the abnormality or its symptom is listed as "Operation Frequency". 【0056】 Process mining As an example, let's describe the case where log analysis result data is created using process mining. In this case, the log analysis unit 113 can create log analysis result data by following the steps 2-1 to 2-3 below. 【0057】 Procedure 2-1: The log analysis unit 113 extracts operation log data from the operation log data acquired in step S105 above, in the same manner as in procedure 1-1 above, from the period from the occurrence of an abnormality or its indication until the end of the abnormality or its indication. 【0058】 Step 2-2: The log analysis unit 113 uses the operation log data extracted in Step 2-1 to calculate the names of the operations performed between the occurrence of the anomaly or its indication and the termination of the anomaly or its indication, as well as the order of those operations. This creates a corresponding operation procedure that represents the names of the operations performed between the occurrence of the anomaly or its indication and the termination of the anomaly or its indication, and the order of those operations. The log analysis unit 113 can also create the corresponding operation procedure using process mining. 【0059】 Step 2-3: The log analysis unit 113 assigns a data name containing the name of the anomaly or its symptom to the corresponding operation procedure created in Step 2-2 above. For example, the log analysis unit 113 can create the data name by setting the name of the anomaly or its symptom to the variable part {} of the pre-created template "{} Operation procedure when an anomaly occurs" (where {} is the variable part). As a result, the corresponding operation procedure with a data name containing the name of the anomaly or its symptom is obtained as log analysis result data. 【0060】 As an example, Figure 8 shows the response procedure when the response procedure is created as a JPEG image file in step 2-2 above, and the file name "Response Procedure for Abnormal Lower Limit Occurrence of Internal Gas Temperature" is created as the data name in step 2-3 above. In the response procedure 1300 shown in Figure 8, it indicates that the abnormality was recovered after operation A, operation A → operation B → operation C, or operation A → operation B was performed after the abnormal lower limit of the internal gas temperature occurred. 【0061】 Following step S106 above, the log analysis unit 113 of the operation procedure DB construction device 110 stores the log analysis result data created in step S106 above into the log analysis result DB 115 (step S107). 【0062】 The text conversion unit 112 of the operation procedure DB construction device 110 creates vector data by vectorizing the text in the log analysis result data represented by the log analysis result data created in step S106 above (step S108). That is, the text conversion unit 112 converts the text data representing the text in the log analysis result into vector data, similar to step S102 above. 【0063】 Here, for example, if the log analysis result data is an image file, the text conversion unit 112 should convert the text data representing the data name of the log analysis result data into vector data. For example, if the corresponding operation list 1100 shown in Figure 6 is created in step S106 above, the text conversion unit 112 should convert the text data representing its data name "Corresponding operation list when an abnormality occurs at the lower limit of the furnace gas temperature" into vector data. Similarly, for example, if the corresponding operation procedure 1300 shown in Figure 8 is created in step S106 above, the text conversion unit 112 should convert the text data representing its data name "Corresponding operation procedure when an abnormality occurs at the lower limit of the furnace gas temperature" into vector data. 【0064】 On the other hand, if, for example, the log analysis result data is a text file, the text conversion unit 112 only needs to convert the text data representing the text contained in the text file into vector data. For example, if the supplementary information 1200 shown in Figure 7 is created in step S106 above, the text conversion unit 112 only needs to convert the text data representing the text contained in the supplementary information 1200 into vector data. 【0065】 The text conversion unit 112 of the operation procedure DB construction device 110 creates operation procedure data by associating the vector data created in step S108 with the text data from which the vector data was converted and the storage location of the log analysis result data corresponding to the text data, and then stores the operation procedure data in the operation procedure DB 114 (step S109). Note that the log analysis result data corresponding to the text data is, if the log analysis result data is an image file, the log analysis result data in which the text represented by the text data is included in the data name, and if the log analysis result data is a text file, the log analysis result data in which the text represented by the text data is included. 【0066】 <Example of abnormality monitoring process> Figure 9 is a flowchart showing an example of the abnormality monitoring process according to this embodiment. Steps S201 to S207 below are repeatedly executed at predetermined time intervals (e.g., the measurement cycle of measured values). 【0067】 The abnormality detection unit 121 of the abnormality monitoring device 120 detects the occurrence of an abnormality or its signs by monitoring at least one of the measurement value data stored in the measurement value DB 304 and the operation log data stored in the operation log DB 305 (step S201). The abnormality detection unit 121 may, for example, detect the occurrence of an abnormality or its signs by using at least one of the following abnormality detection methods 1 to 4. Alternatively, the abnormality detection unit 121 may detect the occurrence of an abnormality or its signs by using, for example, a combination of the following abnormality detection methods 1 to 4. 【0068】 • Anomaly detection method 1 The anomaly detection unit 121 determines that an anomaly or its indication has occurred if a certain predetermined measurement value included in the most recent measurement data is not within a predetermined upper or lower threshold, and that no anomaly or its indication has occurred if it is not within the upper or lower threshold. 【0069】 • Anomaly detection method 2 The anomaly detection unit 121 constructs a prediction model, such as an autoregressive model, from the measurement data for the most recent predetermined period, and then uses this prediction model to predict the measurement data after a predetermined time has elapsed (hereinafter also referred to as "predicted measurement data"). The anomaly detection unit 121 then determines that an anomaly has occurred if a certain predetermined measurement value included in the predicted measurement data exceeds a predetermined threshold, and that no anomaly or its indication has occurred otherwise. 【0070】 • Anomaly detection method 3 The anomaly detection unit 121 constructs a normal model from measurement data during normal plant operation using methods such as Multivariate Statistical Process Control (MSPC). The anomaly detection unit 121 then calculates the degree of anomaly using this normal model and the most recent measurement data. If the degree of anomaly exceeds a predetermined threshold, it is determined that an anomaly or its signs have occurred; otherwise, it is determined that no anomaly or its signs have occurred. 【0071】 • Anomaly detection method 4 The anomaly detection unit 121 determines that an anomaly or its indication has occurred if the data item "Alarm" in the most recent operation log data is set to "Occurred," and that no anomaly or its indication has occurred otherwise. 【0072】 The abnormality detection unit 121 of the abnormality monitoring device 120 determines whether or not it detected an abnormality or its precursor in step S201 (step S202). 【0073】 If it is not determined in step S202 above that an abnormality or its precursor has occurred, the abnormality monitoring device 120 terminates the abnormality monitoring process. 【0074】 On the other hand, if it is determined in step S202 above that an abnormality or its precursor has occurred, the prompt creation unit 122 of the abnormality monitoring device 120 creates a prompt to instruct the large-scale language model to generate a response procedure for the abnormality or its precursor (step S203). The prompt creation unit 122 can create a prompt, for example, using a pre-created prompt template and information set in the data item "event" included in the most recent operation log data where the data item "alarm" is set to "occurrence". A prompt template is a template for a prompt. In a prompt template, part of the prompt is a variable, and the prompt is created by setting the information set in the data item "event" included in the operation log data to that variable part. 【0075】 As an example, let's assume the prompt template is as shown in Figure 10. In the prompt template 2100 shown in Figure 10, {} represents the variable part. Also, let's assume that the data item "Event" included in the most recent operation log data where the data item "Alarm" is set to "Occurred" has "Lower limit of furnace gas temperature" set. In this case, prompt 2200 shown in Figure 11 is created. In prompt 2200 shown in Figure 11, "Lower limit of furnace gas temperature" is set for the variable part of prompt template 2100 shown in Figure 10. 【0076】 The text search unit 131 of the operation guidance device 130 creates vector data by vectorizing the text contained in the prompt created in step S203 (step S204). That is, the text search unit 131 converts the prompt created in step S203 into vector data. At this time, the text search unit 131 converts the prompt into vector data using the same method as the text conversion unit 112 of the operation procedure DB construction device 110. Alternatively, the text search unit 131 may call the text conversion unit 112 of the operation procedure DB construction device 110 and have the text conversion unit 112 convert the prompt into vector data. Hereafter, for convenience, the vector data obtained by vectorizing the text contained in the prompt will be referred to as "prompt vector data". 【0077】 The text search unit 131 of the operation guidance device 130 searches for operation procedure data stored in the operation procedure DB 114 using the prompt vector data created in step S204 (step S205). For example, the text search unit 131 should search for operation procedure data in the operation procedure DB 114 that contains the vector data with the highest similarity to the prompt vector data. As the similarity between vector data, for example, the cosine similarity between the vectors represented by the vector data can be used. 【0078】 The text generation unit 132 of the operation guidance device 130 generates a response procedure for the abnormality or its signs using the prompt created in step S203 and the text data included in the operation procedure data retrieved in step S205 (step S206). For example, the text generation unit 132 generates a response procedure for the abnormality or its signs by inputting the text represented by the text data as prior knowledge, the prompt, and the prior knowledge into a large-scale language model. The text generation unit 132 may use the large-scale language model of the operation guidance device 130, or it may use a large-scale language model provided as an external service. 【0079】 The text generation unit 132 of the operation guidance device 130 uses the data stored in the storage location included in the operation procedure data retrieved in step S205 as reference material and transmits an abnormality notification indicating the occurrence of the abnormality or its signs, the response procedure generated in step S206, and the reference material to the operator device 20 (step S207). As a result, the operator device 20 is presented with the abnormality notification, the response procedure, and the reference material. The reference material is a reference for the response procedure, but since the response procedure is information generated from the text contained in the reference material or its data name, the reference material can also be said to be the basis for the response procedure. 【0080】 Here, Figure 12 shows an example of a display screen that appears on the operator device 20 when an abnormality notification, response procedures, and reference materials are presented. The display screen 3100 shown in Figure 12 includes an abnormality notification display area 3110 that displays the content of the abnormality notification, a response procedure display area 3120 that displays the content of the response procedures, and a reference material display area 3130 that displays the content of the reference materials. This allows the operator to know the content of the abnormality or its signs, the content of the response procedures, and the content of the reference materials. Therefore, by referring to this information, it becomes possible to take a quick and appropriate response to the abnormality or its signs. 【0081】 Furthermore, the display screen 3100 shown in Figure 12 includes an inquiry input field 3140 for inputting inquiries to the operation guidance device 130. For example, if the operator wants to know more detailed information about the nature of an abnormality or its signs, or more detailed information about the response procedure, they can input the desired inquiry written in natural language into the inquiry input field 3140. When an inquiry is entered into the inquiry input field 3140, the inquiry is transmitted from the operator device 20 to the support information display system 10. 【0082】 <Example of inquiry handling process> Figure 13 is a flowchart showing an example of the inquiry handling process according to this embodiment. Steps S301 to S305 below are executed each time an inquiry is entered into the inquiry input field 3140 of the display screen 3100 shown in Figure 12. 【0083】 When the text search unit 131 of the operation guidance device 130 receives an inquiry from the operator device 20, it acquires this inquiry as a prompt (step S301). 【0084】 The text search unit 131 of the operation guidance device 130 creates vector data by vectorizing the text contained in the prompt acquired in step S301 (step S302). That is, the text search unit 131 converts the prompt acquired in step S301 into vector data. At this time, the text search unit 131 converts the prompt into vector data using the same method as the text conversion unit 112 of the operation procedure DB construction device 110. Alternatively, the text search unit 131 may call the text conversion unit 112 of the operation procedure DB construction device 110 and have the text conversion unit 112 convert the prompt into vector data. Hereafter, for convenience, the vector data obtained by vectorizing the text contained in the prompt will be referred to as "prompt vector data". 【0085】 The text search unit 131 of the operation guidance device 130 searches for operation procedure data stored in the operation procedure DB 114 using the prompt vector data created in step S303 (step S303). For example, the text search unit 131 should search for operation procedure data in the operation procedure DB 114 that contains the vector data with the highest similarity to the prompt vector data. As the similarity between vector data, for example, the cosine similarity between the vectors represented by the vector data can be used. 【0086】 The text generation unit 132 of the operation guidance device 130 generates an answer to the query using the prompt obtained in step S301 and the text data included in the operation procedure data retrieved in step S303 (step S304). For example, the text generation unit 132 uses the text represented by the text data as prior knowledge and generates an answer to the query by inputting the prompt and the prior knowledge into a large-scale language model. The text generation unit 132 may use the large-scale language model of the operation guidance device 130, or it may use a large-scale language model provided as an external service. 【0087】 The text generation unit 132 of the operation guidance device 130 uses the data stored in the storage location included in the operation procedure data retrieved in step S303 as reference material, and transmits the answer and the reference material to the operator device 20 (step S305). As a result, the operator device 20 is presented with the answer to the inquiry and the reference material. 【0088】 Here, Figure 14 shows an example of a display screen shown on the operator device 20 when an answer to an inquiry and reference materials are provided. The display screen 3200 shown in Figure 14 includes an inquiry display field 3210 that displays the content of the inquiry, an answer display field 3220 that displays the answer to the inquiry, and a reference material display field 3230 that displays the content of the reference materials. This allows the operator to know the content of the answer to the inquiry and the content of the reference materials. Therefore, by referring to this information, it becomes possible to take a quick and appropriate response to, for example, an abnormality or its signs. Although the reference materials are materials that serve as reference for the answer, the answer is information generated from the text contained in the reference materials or their data names, so the reference materials can also be said to be the materials that form the basis of the answer. 【0089】 Furthermore, the display screen 3200 shown in Figure 14 includes an inquiry input field 3240 for entering inquiries to the operation guidance device 130. For example, if the operator wants to know more detailed information about the content of the answer, they can enter the desired inquiry written in natural language into the inquiry input field 3240. When an inquiry is entered into the inquiry input field 3240, that inquiry is transmitted from the operator device 20 to the support information display system 10. 【0090】 <Example Hardware Configuration> The operation procedure database construction device 110, the abnormality monitoring device 120, the operation guidance device 130, the operator device 20, and the data acquisition device 303 can be realized, for example, by the hardware configuration of the computer 500 shown in Figure 15. Figure 15 is a diagram showing an example of the hardware configuration of the computer 500. 【0091】 The computer 500 shown in Figure 15 includes an input device 501, a display device 502, an external interface 503, a communication interface 504, a RAM (Random Access Memory) 505, a ROM (Read Only Memory) 506, an auxiliary storage device 507, and a processor 508. Each of these hardware components is connected to the others via a bus 509 for communication. 【0092】 The input device 501 is, for example, a keyboard, mouse, touch panel, physical buttons, etc. The display device 502 is, for example, a display, display panel, etc. Note that the computer 500 does not necessarily have to have at least one of the input device 501 and the display device 502. 【0093】 External I / F 503 is an interface with external devices such as recording media 503a. Examples of recording media 503a include CD (Compact Disc), DVD (Digital Versatile Disk), SD memory card (Secure Digital memory card), and USB (Universal Serial Bus) memory card. 【0094】 The communication interface 504 is an interface for connecting to a communication network. The RAM 505 is a volatile semiconductor memory (storage device) that temporarily holds programs and data. The ROM 506 is a non-volatile semiconductor memory (storage device) that can retain programs and data even when the power is turned off. The auxiliary storage device 507 is a non-volatile storage device such as an HDD, SSD, or flash memory. The processor 508 is an arithmetic unit such as a CPU or GPU (Graphics Processing Unit). 【0095】 Note that the hardware configuration of computer 500 shown in Figure 15 is just one example, and the hardware configuration of computer 500 is not limited to this. For example, computer 500 may have multiple auxiliary storage devices 507 and multiple processors 508, it may not have some of the hardware shown, or it may have various other hardware components besides the hardware shown. 【0096】 <Summary> As described above, the operation support system 1 according to this embodiment creates operation procedure data representing the operating procedures when an abnormality or its signs occur, based on technical documents such as operation manuals and training materials used when operating the plant. For this reason, the operation support system 1 according to this embodiment can create operation procedure data without necessarily requiring past operating performance data. 【0097】 Furthermore, in the driver assistance system 1 according to this embodiment, when an abnormality or its signs occur, the system searches for operation procedure data related to the abnormality or its signs, and then generates a response procedure using a large-scale language model based on this operation procedure data. Therefore, the system can present the driver with a response procedure for the abnormality or its signs without requiring any input from the driver. 【0098】 The present invention is not limited to the embodiments specifically disclosed above, and various modifications, changes, and combinations with known technologies are possible as long as they do not deviate from the spirit described in the claims. [Explanation of Symbols] 【0099】 1. Driver assistance system 10. Support Information Presentation System 20 Operator's equipment 30 Plant Systems 110 Operating Procedure DB Construction Device 111 Text acquisition unit 112 Text Conversion Unit 113 Log Analysis Department 114 Operating Procedure DB 115 Log Analysis Results DB 120 Abnormality monitoring device 121 Anomaly detection unit 122 Prompt Creation Section 130 Operation Guidance Device 131 Text Search Section 132 Text generation unit 301 Plant Equipment 302 Sensor 303 Data Acquisition Device 304 Measurement Value Database 305 Operation Log DB 400 Technical Documents 500 Computers 501 Input device 502 Display device 503 External I / F 503a Recording medium 504 Communication I / F 505 RAM 506 ROM 507 Auxiliary storage 508 Processors Bus 509

Claims

[Claim 1] A construction unit constructs a database containing data that includes text information representing the content of the documents and storage location information indicating where the documents are stored, based on documents related to the target operation. When a specific event occurs in the aforementioned target, a generation unit generates a response procedure for the event based on the text information contained in the data corresponding to the event among the data constituting the database, A presentation unit that presents to the user operating the aforementioned target, information indicating the occurrence of the aforementioned event, the aforementioned response procedure, and materials stored in the storage location indicated by the storage location information included in the aforementioned data, An information display device having the following features. [Claim 2] The generating unit is The information presentation device according to claim 1, which generates the response procedure by inputting a first prompt for instructing a large-scale language model to generate a response procedure when the aforementioned event occurs, and the text information, into the large-scale language model. [Claim 3] The data further includes first vector information obtained by vectorizing the text information, The generating unit is The database is searched for the data with the highest similarity between the second vector information obtained by vectorizing the first prompt and the first vector information. The information presentation device according to claim 2, which generates the corresponding procedure by inputting the first prompt and the text information contained in the retrieved data into the large-scale language model. [Claim 4] The aforementioned construction unit is The information presentation device according to claim 1, which constructs a database containing data that includes text information representing the content of the results of a predetermined analysis performed on the past operating performance data of the subject, and storage location information indicating the storage location of the results. [Claim 5] The generating unit is The information presentation device according to claim 2, further generating a response to the inquiry by inputting a second prompt for instructing the large-scale language model to generate a response to the inquiry from the user, and the text information, into the large-scale language model. [Claim 6] The information display device according to any one of claims 1 to 5, wherein the aforementioned materials include an operating manual for operating the subject and educational materials for the user operating the subject. [Claim 7] A construction step involves constructing a database containing data that includes text information representing the content of the documents and storage location information indicating where the documents are stored, based on the documents relating to the target operation. When a specific event occurs in the aforementioned target, a generation step is made to generate a response procedure for the event based on the text information contained in the data corresponding to the event among the data constituting the database. A presentation step in which information representing the occurrence of the event, the response procedure, and the materials stored in the storage location indicated by the storage location information included in the data are presented to the user who drives the aforementioned target, A method of presenting information that a computer performs. [Claim 8] A construction step involves constructing a database containing data that includes text information representing the content of the documents and storage location information indicating where the documents are stored, based on the documents relating to the target operation. When a specific event occurs in the aforementioned target, a generation step is made to generate a response procedure for the event based on the text information contained in the data corresponding to the event among the data constituting the database. A presentation step in which information representing the occurrence of the event, the response procedure, and the materials stored in the storage location indicated by the storage location information included in the data are presented to the user who drives the aforementioned target, A program that causes a computer to execute something.

Images