TW202422257A - Determination device, determination method, program, and plant system - Google Patents

Abstract

[Topic] Provide a determination device, determination method, program for the determination device, and factory system capable of accurately determining abnormalities in a factory. [Solution] The determination device (10) comprises: a data acquisition unit (110) for acquiring data from a plurality of sensors installed in the factory (11); and a determination unit (130) for determining the presence or absence of abnormalities based on the data acquired by the data acquisition unit (110) and a statistical model (151) representing the correlation between the plurality of data. The plurality of sensors includes a prediction target sensor. The determination unit (130) calculates a predicted value of the data obtained from the prediction target sensor using data obtained from a sensor among a plurality of sensors that is installed in the same system as the prediction target sensor and measures the same type of physical quantity as the prediction target sensor, and determines the presence or absence of an abnormality based on the predicted value and the measured value of the data obtained from the prediction target sensor.

Description

Determination device, determination method, program and factory system

The present invention relates to a determination device, a determination method, a program for the determination device and a factory system for determining abnormalities in a factory.

For example, in power plants and incineration plants, sensors are used to obtain data such as temperature and pressure of various operating parts, and the data are always monitored to see if they are normal values. If some data become abnormal values, the cause of the abnormality is estimated based on the location of the corresponding sensor, and appropriate measures are taken to restore it to normal.

When determining whether the acquired specific data is a normal value, for example, as described in the following patent document 1, the predicted value for the data is calculated according to the situation, and it is determined how much the actual measured value of the data deviates from the calculated predicted value. In addition, the "predicted value" mentioned here refers to the value predicted to be represented by the above-mentioned specific data under the assumption that the factory as a whole operates normally. [Prior art document] [Patent document]

[Patent Document 1] Japanese Patent Application Publication No. 64-21509

[The problem the invention is trying to solve]

The above-mentioned predicted value of each data can be calculated, for example, using the measured values of other multiple data that affect the data and a statistical model. The statistical model is a model (for example, a mathematical formula) that expresses the correlation between multiple data, and is prepared based on each data obtained in advance under normal circumstances.

However, the types and quantities of the "other multiple data that affect the data" mentioned above are huge, and if the scale of the factory increases, such data will become even larger. Therefore, most of the statistical models required for the calculation of the predicted value are very complicated, and in order to obtain an accurate statistical model, it is necessary to use a large number of sample data for learning in advance.

However, the time available for learning is limited, and in many cases it is difficult to use a sufficient number of sample data for learning. If the statistical model is learned with a small number of sample data, the calculation accuracy of the predicted value will become low, so it will be difficult to determine the presence or absence of abnormalities when compared with the measured values. In other words, it is difficult to observe the difference between normal and abnormal times for each measured data.

The purpose of the present invention is to provide a determination device, determination method, program for the determination device and factory system that can accurately determine abnormalities in a factory. [Technical means for solving the problem]

The determination device of the present invention is a determination device for determining abnormalities in a factory, and comprises: a data acquisition unit that acquires data from a plurality of sensors installed in the factory; and a determination unit that determines the presence or absence of abnormalities based on the data acquired by the data acquisition unit and a statistical model representing the correlation between the plurality of data. The plurality of sensors installed in the factory include a prediction target sensor. The determination unit of the determination device calculates a predicted value of the data acquired from the prediction target sensor using data acquired from a sensor among the plurality of sensors that is installed in the same system as the prediction target sensor and measures the same type of physical quantity as the prediction target sensor, and determines the presence or absence of abnormalities based on the predicted value and the measured value of the data acquired from the prediction target sensor.

In a determination device having such a configuration, a predicted value of data obtained by the prediction target sensor is calculated using data obtained from a sensor that is installed in the same system as the prediction target sensor and measures the same type of physical quantity as the prediction target sensor. In this case, the statistical model used in the calculation of the predicted value represents the relationship between data consisting of the same system and the same type of physical quantity, so it is easier to express it in a mathematical expression than in the case of including multiple types of physical quantities. Therefore, even if the amount of data used to learn the statistical model is small, the predicted value can be calculated with higher accuracy than in the past. In the event of an abnormality, the difference between the actual measured value obtained by the predicted sensor and the predicted value calculated as the normal value becomes significant, so the presence or absence of an abnormality can be accurately determined based on the two. [Effect of the invention]

According to the present invention, a determination device, a determination method, a program for the determination device and a factory system are provided which can accurately determine abnormalities in a factory.

Hereinafter, the present embodiment will be described with reference to the drawings. In order to facilitate the understanding of the description, the same components are marked with the same symbols as much as possible in each drawing, and repeated description is omitted.

The determination device 10 of the present embodiment constitutes a part of the plant system PS, and is a device for determining an abnormality in the plant 11 described later. The "abnormality" in the plant 11 includes, for example, a state where the plant 11 cannot operate due to a physical failure such as a pipe burst, or a state where there is a high possibility that the plant 11 will stop. In addition, for example, the operation efficiency of the plant 11 is significantly reduced due to some reason, and the situation that is not accompanied by a physical failure and deviates from a good operation state is also included in the above-mentioned "abnormality".

Before explaining the determination device 10, the configuration of the plant system PS will be mainly explained with reference to Fig. 1. As shown in Fig. 1, the plant system PS includes a plant 11, a DCS 12, a determination device 10, and a monitoring terminal 20.

The plant 11 of the present embodiment is a power plant including a boiler (not shown). In the boiler, water vapor is generated by the heat of high-temperature exhaust gas, and the turbine of the generator is driven by the water vapor. When the plant 11 is in operation, the data (temperature or pressure, etc.) of each part in the exhaust system or water system is always monitored, and control is performed so that each data is maintained at a normal value. Furthermore, the plant 11 can be a power plant as in the present embodiment, or it can be another type of plant. For example, it can be an incineration plant or a chemical plant.

The DCS 12 is also called a "distributed control system" and is a device for controlling the factory 11. The DCS 12 controls a part of the factory 11 (for example, a boiler), but can also control the entire factory 11.

DCS12 sends control signals to each of the various machines (e.g., blowers) installed in the factory 11 to control the operation of the factory 11. In addition, DCS12 receives signals from each of the multiple sensors installed in various parts of the factory 11 to obtain the status of the factory 11. In the table shown in FIG2, multiple sensors installed in the factory 11 are summarized for each system. The "system" mentioned here is a system that constitutes a series of paths for the flow of the same type of media, such as an exhaust system, a water system, an electric power system, etc.

In the present embodiment, there are three systems in the factory 11, which are composed of system A, system B and system C, and a plurality of sensors are installed in each system. In the example of FIG2 , a total of eight sensors, S11 to S18, are installed in various parts of system A. Among them, S11, 12, and 13 are sensors for measuring pressure, S14, 15, and 16 are sensors for measuring temperature, and S17 and 16 are sensors for measuring flow. The number or types of sensors (S21, etc.) shown in system B and system C shown in FIG2 are as shown in FIG2 . Furthermore, the number of sensors actually installed in each system is more than the number shown in FIG2 .

As described above, the determination device 10 is a device for determining abnormalities in the factory 11. The determination device 10 is configured as a computer system having a CPU, ROM, RAM, etc. (not shown), and is, for example, installed in the same place as the factory 11. The determination device 10 may also be configured as a cloud server located at a different location from the factory 11.

The determination device 10 can perform two-way communication with the DCS 12. The determination device 10 obtains each measured value of each sensor installed in the factory 11 by communicating through the DCS 12. Instead of this, the determination device 10 can also directly receive signals from each sensor.

The monitoring terminal 20 is a terminal device provided as an interface between users. By operating the monitoring terminal 20, the user can input information required for the action setting of the determination device 10, etc. In addition, the monitoring terminal 20 can report the determination result based on the determination device 10 to the user by displaying various information on the screen 200. The monitoring terminal 20 can be a fixed terminal provided at a specific location, or a mobile communication terminal that can be carried by the user. The monitoring terminal 20 can be regarded as a part of the determination device 10.

The structure of the determination device 10 will be described with reference to Fig. 1. The determination device 10 includes a data acquisition unit 110, a selection unit 120, a determination unit 130, a notification unit 140, and a storage unit 150 as functional elements.

The data acquisition unit 110 is a part that processes data acquired from a plurality of sensors installed in the factory 11. The data acquisition unit 110 repeatedly acquires data from all sensors installed in the factory 11 through communication with the DCS 12.

The selection unit 120 is a part that performs processing to select data to be used in the processing of the determination unit 130 from the plurality of data acquired by the data acquisition unit 110. The specific content of the processing performed by the selection unit 120 will be described later.

The determination unit 130 is a part that performs the following processing: based on the data obtained by the data acquisition unit 110 (specifically, the data selected by the selection unit 120) and the statistical model, it determines whether there is an abnormality in the factory 11. "Statistical model" is a model that represents the correlation between multiple data obtained by each sensor, for example, expressed as a mathematical expression. The statistical model is pre-made based on the data obtained from each sensor when the factory 11 is operating normally as a whole, and is stored in the storage unit 150 described later. The statistical model can be learned and updated each time during the normal operation of the factory 11.

For example, the determination unit 130 can input data measured by a plurality of sensors into a statistical model, and calculate a predicted value of data obtained by a specific sensor as its output. The "predicted value" calculated in this way is predicted to be the value indicated by the specific sensor under the assumption that the factory 11 is operating normally as a whole. Therefore, when the measured value actually obtained by the sensor deviates significantly from the calculated predicted value, the determination unit 130 can determine that an abnormality has occurred in the factory 11.

The notification unit 140 is a part that performs processing to notify the user based on the determination result of the determination unit 130. The notification unit 140 notifies the user of the determination result by displaying it on the screen 200 of the monitoring terminal 20. The specific aspects of the notification by the notification unit 140 will be described later.

The memory unit 150 is a non-volatile memory device provided in the determination device 10, for example, a HDD or an SSD. The memory unit 150 may be a file server provided in a location different from the determination device 10. The memory unit 150 stores various information required for the processing performed by the determination device 10 including the statistical model described above.

As described later, a statistical model is individually created for each sensor installed in the factory 11, and each statistical model is stored in the memory unit 150. Hereinafter, each statistical model stored in the memory unit 150 is also referred to as a "statistical model 151".

The outline of the processing executed by the determination device 10 will be described with reference to Fig. 3. Fig. 3 schematically shows the flow of information in each unit of the determination device 10.

Hereinafter, a specific sensor used for abnormality determination among the plurality of sensors installed in the factory 11 is also referred to as a "prediction target sensor". The determination unit 130 selects any one sensor from the plurality of sensors as a prediction target sensor, and calculates the above-mentioned "prediction value" in the prediction target sensor. Thereafter, the actual measurement value and the prediction value in the prediction target sensor are compared to determine the presence or absence of abnormality.

The data acquisition unit 110 acquires data from all sensors including the prediction target sensor, and sends the data to the selection unit 120. The selection unit 120 selects data obtained from sensors that are installed in the same system as the prediction target sensor and measure the same type of physical quantity as the prediction target sensor from among these data. For example, in the case where S14 in FIG. 2 becomes the prediction target sensor, the selection unit 120 selects sensors S15 and S16 that are installed in the same system A and measure the same "temperature" as S14. The selection unit 120 sends the data thus selected to the prediction value calculation unit 131 that is a part of the determination unit 130. In FIG. 3, the data thus sent is represented as "d2".

The selection unit 120 sends the data of the actual measurement value obtained by the prediction target sensor among the data sent from the data acquisition unit 110 to the abnormality calculation unit 132 which is a part of the determination unit 130. The data sent in this way is indicated as "d1" in FIG3.

The predicted value calculation unit 131 reads out the statistical model 151 corresponding to the prediction target sensor from the plurality of statistical models 151 stored in the memory unit 150. The statistical model 151 is prepared in advance and represents the correlation between the data obtained from the sensor that is installed in the same system as the prediction target sensor and measures the same type of physical quantity as the prediction target sensor (i.e., d2 in FIG. 3 ) and the data measured by the prediction target sensor.

The predicted value calculation unit 131 calculates the predicted value of the data measured by the prediction target sensor based on the statistical model 151 and the data (d2) selected by the selection unit 120, and sends the predicted value to the abnormality calculation unit 132. In FIG3, the predicted value sent in this way is represented as "d1'".

As described above, both the actual value d1 of the data obtained by the prediction target sensor and the predicted value d1' of the data measured by the prediction target sensor are sent to the abnormality calculation unit 132.

The predicted value d1' is calculated under the premise that the entire factory 11 operates normally. Therefore, when no abnormality occurs in the factory 11, the measured value d1 becomes a value substantially equal to the predicted value d1', and the difference between the two becomes smaller. On the other hand, when an abnormality occurs in the factory 11, the measured value d1 becomes a value deviated from the predicted value d1', and the difference between the two becomes larger. Therefore, the difference between the measured value d1 and the predicted value d1' indicates the degree of the abnormality occurring in the factory 11.

The anomaly calculation unit 132 calculates the "abnormality" which is an indicator indicating the degree of abnormality occurring in the factory 11. In the present embodiment, the anomaly is calculated as a value obtained by normalizing the difference between the measured value d1 and the predicted value d1'. For example, the anomaly is calculated as a value in the range of 0 to 1. In this case, when the measured value d1 and the predicted value d1' are consistent, the anomaly becomes 0. In addition, as the difference between the measured value d1 and the predicted value d1' increases, the anomaly increases, and after the difference increases to a certain extent, the anomaly is calculated as 1.

Considering the sign of the difference between the measured value d1 and the predicted value d1', the anomaly can be calculated as a value from -1 to 1. In this case, when the measured value d1 is significantly larger than the predicted value d1', the anomaly becomes 1, and when the measured value d1 is significantly smaller than the predicted value d1', the anomaly becomes -1.

Furthermore, the abnormality calculated by the abnormality calculation unit 132 may be a value calculated by a method different from the above as long as it can be used as an indicator of the degree of abnormality occurring in the factory 11. For example, the difference between the measured value d1 and the predicted value d1' may be directly used as the abnormality.

The abnormality calculated by the abnormality calculation unit 132 is sent to the notification unit 140. The notification unit 140 notifies the user by displaying the abnormality on the screen 200 of the monitoring terminal 20.

As described above, the determination unit 130 of the determination device 10 of the present embodiment is configured to calculate a predicted value (d1') of the data obtained from the prediction object sensor using data obtained from a sensor among a plurality of sensors installed in the factory 11, which is installed in the same system as the prediction object sensor and measures the same type of physical quantity as the prediction object sensor, and determine the presence or absence of an abnormality based on the predicted value (d1') and the measured value (d1) of the data obtained from the prediction object sensor.

The statistical model 151 used in the calculation of the predicted value is prepared as a model that expresses the relationship between data composed of the same system and the same type of physical quantity as the prediction target sensor. That is, the statistical model 151 is prepared as a model that does not include data measured by a sensor installed in a system different from the prediction target sensor, or data of a physical quantity of a different type from the data measured by the prediction target sensor. As a result, the statistical model 151 of the present embodiment is easier to express in a mathematical expression or the like than in the case of including a plurality of types of physical quantities, and thus becomes a relatively simple model.

As a result, even when learning a statistical model with a small amount of data, the predicted value calculation unit 131 can calculate the predicted value with higher accuracy than before. When an abnormality occurs, the difference between the predicted value d1' and the measured value d1 is significant, so the determination unit 130 can accurately determine the presence or absence of an abnormality based on the comparison between the two.

Furthermore, the data d2 selected by the selection unit 120 and sent to the prediction value calculation unit 131 may also include data of actual values obtained from the prediction target sensor. For example, in the configuration in which the statistical model 151 is updated each time by deep learning, it is preferable to calculate the prediction value based on the statistical model in the form of data based on actual values of the prediction target sensor.

Therefore, it can be seen that the determination unit 130 of the present embodiment is able to calculate the predicted value of the data obtained from the prediction object sensor using only the data obtained from a sensor among a plurality of sensors that is installed in the same system as the prediction object sensor and measures the same type of physical quantity as the prediction object sensor, or using the data obtained by adding the data measured by the prediction object sensor to the data, and determine the presence or absence of an abnormality based on the predicted value and the measured value of the data obtained from the prediction object sensor.

The specific flow of the processing performed by the determination device 10 will be described with reference to the flowchart of Fig. 4. The series of processing shown in Fig. 4 is repeatedly performed by the determination device 10 every predetermined period during the operation of the plant 11.

In the first step ST1, the data acquisition unit 110 performs a process of acquiring data from all sensors installed in the factory 11. The process performed in step ST1 corresponds to the "acquisition step" and "data acquisition process" in this embodiment.

In step ST2 after step ST1, for example, the determination unit 130 determines one prediction target sensor from the sensors installed in the factory 11. In this embodiment, each of the plurality of sensors installed in the factory 11 is sequentially determined as a prediction target sensor, and the presence or absence of abnormality is determined based on the comparison between the measured value and the predicted value of each prediction target sensor. The order in which each sensor is determined as a prediction target sensor can be arbitrarily set.

In step ST3 following step ST2, the selection unit 120 processes data obtained from a sensor that is installed in the same system as the prediction target sensor and measures the same type of physical quantity as the prediction target sensor, that is, selects data for calculating the prediction value.

In step ST4 following step ST3, the predicted value calculation unit 131 of Fig. 3 reads the statistical model 151 corresponding to the prediction target sensor from the memory unit 150. In the subsequent step ST5, the predicted value calculation unit 131 calculates the predicted value of the data measured by the prediction target sensor based on the statistical model 151 read in step ST4 and the data for prediction value calculation selected in step ST3 (d2 in Fig. 3).

In step ST6 after step ST5, the abnormality calculation unit 132 of FIG3 performs a process of calculating the abnormality. As described above, the abnormality calculation unit 132 calculates the abnormality based on the data of the actual value obtained by the prediction target sensor (d1 in FIG3) and the predicted value (d1' in FIG3) calculated in step ST5. The process performed in step ST6 corresponds to the "determination process" and "determination process" in this embodiment. In step ST7 after step ST6, the notification unit 140 performs a process of displaying the abnormality on the screen 200 of the monitoring terminal 20.

In step ST8 after step ST7, it is determined whether the calculation and notification of the abnormality have been completed for all sensors installed in the factory 11. If it has been completed for all sensors, the series of processing shown in Figure 4 is terminated. If there are sensors that have not been selected as prediction target sensors, the process proceeds to step ST9. In step ST9, one of the sensors that have not been selected as prediction target sensors is selected and the process of determining it as the next prediction target sensor is performed. Thereafter, the process after step ST3 is executed again.

As a result of repeating such processing, the determination unit 130 determines the presence or absence of an abnormality when the sensor is set as a prediction target sensor for each of the plurality of sensors installed in the factory 11. Furthermore, the notification unit 140 individually reports the determination result corresponding to each of the plurality of sensors. That is, the degree of abnormality when the sensor is set as a prediction target sensor is calculated and reported for each of the plurality of sensors. Furthermore, the sensors set as prediction target sensors may be all sensors as in the present embodiment, or may be only sensors installed in a part of the factory 11.

An example of the notification by the notification unit 140 will be described with reference to Fig. 5. When the notification is performed, time series graphs 201, 202, 203 and an abnormality distribution graph 210 are displayed on the screen 200 of the monitoring terminal 20.

The time series graphs 202 and 203 are graphs that respectively display the change of the abnormality calculated for the sensors S14 and S15 in Fig. 2 in time series. The time series graph 201 is a graph that displays the difference (temperature difference) between the data measured by the sensors S14 and S15 in time series.

Furthermore, the sensor to be displayed in the time series graph may be other than the above-mentioned sensors. The data to be displayed in the time series graph may be arbitrarily set by the user through the operation of the monitoring terminal 20.

The anomaly distribution diagram 210 displays a plurality of icons 211 representing sensors installed in various parts on a configuration diagram representing the entire factory 11. The positions where the icons 211 are displayed correspond to the positions where the sensors are installed in the factory 11. Furthermore, the color of the icon 211 represents the value of the anomaly calculated by the determination unit 130 for the corresponding sensor. For example, the icon 211 is displayed in a color selected from a gradient color in which the closer the anomaly calculated for the corresponding sensor is to 0, the closer it is to blue, and the closer the anomaly is to 1 or -1, the closer it is to red. The color of each icon 211 is updated in the processing of the subsequent step ST7 according to the anomaly calculated in step ST6 of FIG. 4. That is, the notification unit 140 of the present embodiment individually notifies the determination results corresponding to each sensor by changing the color of each icon 211 displayed on the abnormality distribution diagram 210.

Furthermore, the icon 211 displayed on the anomaly distribution diagram 210 can be displayed only for the sensor corresponding to the data selected by the selection unit 120 (that is, the data used to calculate the predicted value), or can be displayed for all sensors installed in the factory 11.

As described above, the icon 211 may indicate the value of the abnormality by color, but may also indicate the result of determination whether the abnormality exceeds a predetermined threshold by two types of colors. In other words, the notification unit 140 may notify the presence or absence of abnormality in each unit of the factory 11 including the degree (i.e., the abnormality), or may simply notify the presence or absence of abnormality.

The above determination method performed by the determination device 10 is realized by, for example, a program stored in a non-volatile memory device (not shown) provided in the determination device 10. The determination unit 130 of the determination device 10 is operated according to the program, thereby executing the above-mentioned "data acquisition processing" or "determination processing" and the like. A part or all of the program can be sent to the determination device 10 from the outside and temporarily written to the memory device of the determination device 10.

[Note] The following is a list of technical ideas that can be mastered based on the above implementation methods. (Note 1) A determination device for determining abnormalities in a factory, the determination device comprising: a data acquisition unit for acquiring data from a plurality of sensors installed in the factory; and a determination unit for determining the presence or absence of abnormalities based on the data acquired by the data acquisition unit and a statistical model representing the correlation between the plurality of data, wherein the plurality of sensors include a prediction target sensor, the determination unit calculates a predicted value of the data acquired from the prediction target sensor using data acquired from a sensor among the plurality of sensors that is installed in the same system as the prediction target sensor and measures the same type of physical quantity as the prediction target sensor, and determines the presence or absence of abnormalities based on the predicted value and the measured value of the data acquired from the prediction target sensor. (Appendix 2) A determination device as in Appendix 1, further comprising: A selection unit that selects data obtained from a sensor that is installed in the same system as the prediction target sensor and measures the same type of physical quantity as the prediction target sensor from the plurality of data obtained by the data acquisition unit. (Appendix 3) A determination device as in Appendix 1 or 2, wherein the data used to calculate the predicted value also includes data obtained from the prediction target sensor. (Appendix 4) A determination device as in any one of Appendixes 1 to 3, further comprising: A notification unit that notifies the determination result based on the determination unit. (Appendix 5) A determination device as in Appendix 4, wherein the determination unit determines the presence or absence of abnormality when each of the plurality of sensors is set as a prediction target sensor, and the notification unit individually notifies the determination result corresponding to each of the plurality of sensors.

Above, the present embodiment is described with reference to specific examples. However, the present disclosure is not limited to these specific examples. In these specific examples, as long as they have the characteristics of the present disclosure, the parts that are appropriately designed and changed by technical personnel in this field are also included in the scope of the present disclosure. The various requirements and their configurations, conditions, shapes, etc. possessed by the aforementioned specific examples should not be limited to the examples, but can be appropriately changed. The various requirements possessed by the aforementioned specific examples can be appropriately changed in combination as long as they do not cause technical contradictions. This application claims priority based on Japanese Patent Application No. 2022-190180 filed on November 29, 2022. The entire contents of the Japanese application are cited in this specification by reference.

10: Judgment device 11: Factory 110: Data acquisition unit 120: Selection unit 130: Judgment unit 140: Notification unit PS: Factory system

[Figure 1] is a diagram schematically showing the overall structure of a factory system of the implementation method. [Figure 2] is a diagram showing a plurality of sensors installed in the factory, summarized for each system. [Figure 3] is a diagram for explaining the functions of the determination unit, etc., of the determination device. [Figure 4] is a flow chart showing the flow of processing performed by the determination device. [Figure 5] is a diagram showing an example of information displayed on the screen of the monitoring terminal.

10: Judgment device

11: Factory

12:DCS

20: Monitoring terminal

110: Data Acquisition Department

120:Selection Department

130: Judgment Department

140: Report Department

150: Memory Department

151: Statistical Model

200: Screen

PS: Factory system

Claims (8)

A determination device for determining an abnormality in a factory, the determination device comprising: a data acquisition unit for acquiring data from a plurality of sensors installed in the factory; and a determination unit for determining the presence or absence of an abnormality based on the data acquired by the data acquisition unit and a statistical model representing the correlation between the plurality of data, the plurality of sensors including a prediction target sensor, the determination unit calculates a predicted value of the data acquired from the prediction target sensor using data acquired from a sensor among the plurality of sensors that is installed in the same system as the prediction target sensor and measures the same type of physical quantity as the prediction target sensor, and determines the presence or absence of an abnormality based on the predicted value and the measured value of the data acquired from the prediction target sensor. The determination device of claim 1, further comprising: A selection unit, which selects, from the plurality of data acquired by the data acquisition unit, data acquired from the aforementioned sensor that is set in the same system as the aforementioned prediction target sensor and measures the same type of physical quantity as the aforementioned prediction target sensor. As in the determination device of claim 1, wherein, the data used to calculate the aforementioned predicted value also includes data obtained from the aforementioned predicted object sensor. The determination device of claim 1, further comprising: A notification unit for notifying the determination result based on the determination unit. The determination device of claim 4, wherein: the determination unit determines individually for each of the plurality of sensors whether there is an abnormality when the sensor is set as the prediction target sensor; the notification unit individually notifies the determination result corresponding to each of the plurality of sensors. A determination method for determining an abnormality in a factory, the determination method comprising the following steps: an acquisition step of acquiring data from a plurality of sensors installed in the factory; and a determination step of determining the presence or absence of an abnormality based on the data acquired in the acquisition step and a statistical model representing the correlation between the plurality of data, the plurality of sensors including a prediction target sensor, in the determination step, using data acquired from a sensor among the plurality of sensors that is installed in the same system as the prediction target sensor and measures the same type of physical quantity as the prediction target sensor, a predicted value of the data acquired from the prediction target sensor is calculated, and the presence or absence of an abnormality is determined based on the predicted value and the measured value of the data acquired from the prediction target sensor. A program for a determination device, the determination device determines an abnormality in a factory, the program causing the determination device to perform the following processing: Data acquisition processing, obtaining data from a plurality of sensors installed in the factory; and Determination processing, determining the presence or absence of an abnormality based on the data obtained in the data acquisition processing and a statistical model representing the correlation between the plurality of data, The plurality of sensors include a prediction target sensor, The aforementioned program causes the aforementioned determination device to perform the following processing as the aforementioned determination processing: using the data obtained from the aforementioned sensor among the plurality of aforementioned sensors, which is set in the same system as the aforementioned prediction target sensor and measures the same type of physical quantity as the aforementioned prediction target sensor, calculate the predicted value of the data obtained from the aforementioned prediction target sensor, and determine the presence or absence of abnormality based on the predicted value and the actual measured value of the data obtained from the aforementioned prediction target sensor. A factory system includes a factory and a determination device for determining an abnormality in the factory. In the factory system, the determination device comprises: a data acquisition unit for acquiring data from a plurality of sensors installed in the factory; and a determination unit for determining the presence or absence of an abnormality based on the data acquired by the data acquisition unit and a statistical model representing the correlation between the plurality of data. The plurality of sensors include a prediction target sensor. The determination unit calculates a predicted value of the data obtained from the prediction target sensor using data obtained from the aforementioned sensor among the plurality of aforementioned sensors that is installed in the same system as the aforementioned prediction target sensor and measures the same type of physical quantity as the aforementioned prediction target sensor, and determines the presence or absence of abnormality based on the predicted value and the measured value of the data obtained from the aforementioned prediction target sensor.

Images