JP2018139085A - Method, device, system, and program for abnormality prediction - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a method to predict if any abnormality occurs in future.SOLUTION: An abnormality predicting device 10 extracts process data contained in an analysis frame, that is, prescribed time width Wa, from process data which are time-sequence data collected by sensors of a monitored object facility. Then, the abnormality predicting device 10 calculates a frame abnormality evaluation value, that is, an interim evaluation value expressing a probability of abnormality occurrence at a time prescribed period after a prescribed time, by using a learning model predicting abnormality of the monitored object facility whose inputs are the extracted process data. Finally, the abnormality predicting device 10 calculates a predicted abnormality evaluation value which is a final evaluation value expressing the probability of abnormality occurrence at a time prescribed period after a prescribed time, on the basis of the plural calculated frame abnormality evaluation values.SELECTED DRAWING: Figure 1

Description

  The present invention relates to an abnormality prediction method, an abnormality prediction device, an abnormality prediction system, and an abnormality prediction program.

  Conventionally, as a technology related to the prediction of abnormalities using process data from sensors in factories, plants, etc., especially when the definition of sensor abnormal values is not clear and it is difficult to detect abnormalities with thresholds etc. A method for predicting or detecting an abnormality based on process data collected in the past has been proposed.

  Generally, process data expressed as multivariate time-series data is cut out by a window of a predetermined time, and a value related to determination of occurrence of abnormality is calculated by performing an operation. As a specific method, a past model is used to learn a calculation model that expresses the current state or a calculation model that calculates the current (or future) process data itself. If the difference between the evaluation value obtained by applying the calculation model and the actually measured value is large, it is considered that an abnormality has occurred.

  As a method using a calculation model that represents the current state, we define a `` distance '' function that represents the difference between each point in the past data and the currently collected data, and if the distance exceeds the threshold, Define the “subspace” where the data should exist in the normal state using the method to be considered and past data, and determine the anomaly by looking at the separation from the “subspace” calculated from the newly collected data There are ways to do it.

  On the other hand, as a method of creating a calculation model that performs prediction using a supervised learning algorithm, the calculation formula for calculating the current data using the past data for a certain time as an input is a VAR model (vector autoregressive model) There is also a method of discriminating an abnormality by learning using a method, etc., and checking the difference between the predicted value and the actual value.

Japanese Patent Application Laid-Open No. 2006-173782

  As described above, in the conventional method, since an abnormality is detected using currently collected data, there is a problem that it is impossible to predict whether an abnormality will occur in the future. For example, the conventional anomaly detection method uses a past data to learn a calculation model that expresses the current state or a calculation model that calculates the current process data itself. However, there is a problem that it is impossible to predict future abnormalities when the deviation between the evaluation value obtained by applying the calculation model and the actually measured value is large. there were.

  In order to solve the above-described problems and achieve the object, the abnormality prediction method of the present invention is an abnormality prediction method executed by an abnormality prediction device, and includes time-series data collected by a sensor of a monitoring target facility. An extraction process for extracting process data included in a first time width that is a predetermined time width from a certain process data, and the process data extracted by the extraction process are input to predict an abnormality of the monitoring target equipment A frame abnormality evaluation value calculating step for calculating a frame abnormality evaluation value that is an intermediate evaluation value representing the probability of occurrence of an abnormality for a time point after a predetermined time from a predetermined time point using the learning model for Based on the plurality of frame abnormality evaluation values calculated by the frame abnormality evaluation value calculation step, the difference at a time point after a predetermined time from the predetermined time point. There wherein the containing and abnormal predictive evaluation value calculating step of calculating a final evaluation abnormalities predictive evaluation value is representative of the likelihood that occur.

  Further, the abnormality prediction apparatus of the present invention is an extraction unit that extracts process data included in a first time width that is a predetermined time width from process data that is time-series data collected by a sensor of a monitoring target facility. And, using the process data extracted by the extraction means as an input, using a learning model for predicting the abnormality of the monitored equipment, the probability that an abnormality will occur at a time point after a predetermined time from a predetermined time point A frame abnormality evaluation value calculating unit that calculates a frame abnormality evaluation value that is an intermediate evaluation value to be expressed, and a plurality of frame abnormality evaluation values calculated by the frame abnormality evaluation value calculating unit. Abnormal prediction evaluation value calculation that calculates an abnormal prediction evaluation value that is the final evaluation value that represents the probability that an abnormality will occur at the time point And having a stage.

  The abnormality prediction system of the present invention is an abnormality prediction system in which an abnormality prediction device and a parameter learning device are connected via a network, and the abnormality prediction device is collected by a sensor of a monitoring target facility. Process data that is a series of data and abnormality occurrence time data indicating a time when an abnormality has occurred in the monitored facility are transmitted to the parameter learning device, and parameters of a learning model for predicting an abnormality in the monitored facility The communication means received from the parameter learning device, the extraction means for extracting process data included in the first time width which is a predetermined time width from the process data, and the process data extracted by the extraction means are input. Using the learning model, it is confirmed that an abnormality will occur at a time point after a predetermined time from a predetermined time point. A frame abnormality evaluation value calculating unit that calculates a frame abnormality evaluation value that is an intermediate evaluation value representing the likelihood, and the predetermined time point based on a plurality of frame abnormality evaluation values calculated by the frame abnormality evaluation value calculating unit. An abnormality prediction evaluation value calculation means for calculating an abnormality prediction evaluation value that is a final evaluation value representing the probability that an abnormality will occur at a time point after a predetermined time from the above, the parameter learning device, the process data and Learning data receiving means for receiving the abnormality occurrence time data from the abnormality prediction device; learning parameters of the learning model based on the process data and the abnormality occurrence time data received by the learning data receiving means; Learning means for learning, and parameters of the learning model learned by the learning means are used as the abnormality prediction device. And having a parameter transmission means for transmitting to.

  Further, the abnormality prediction program of the present invention is an extraction step for extracting process data included in a first time width that is a predetermined time width from process data that is time-series data collected by a sensor of a monitoring target facility. And, using the process data extracted in the extraction step as an input, using a learning model for predicting an abnormality of the monitored facility, the probability that an abnormality will occur at a time after a predetermined time from a predetermined time A frame abnormality evaluation value calculating step for calculating a frame abnormality evaluation value, which is an intermediate evaluation value to be expressed, and a plurality of frame abnormality evaluation values calculated by the frame abnormality evaluation value calculating step. Calculates the abnormal prediction evaluation value, which is the final evaluation value that represents the probability that an abnormality will occur at a later time point. Characterized in that to execute a that abnormal predicted evaluation value calculation step to the computer.

  According to the present invention, it is possible to predict whether or not an abnormality will occur in the future.

FIG. 1 is a block diagram illustrating a configuration example of the abnormality prediction apparatus according to the first embodiment. FIG. 2 is a diagram illustrating an abnormality prediction evaluation value calculation process executed by the abnormality prediction unit. FIG. 3 is a diagram for explaining the outline of the abnormality prediction process executed by the abnormality prediction device. FIG. 4 is a flowchart illustrating an example of a flow of abnormality prediction processing in the abnormality prediction device according to the first embodiment. FIG. 5 is a flowchart illustrating an example of a learning process flow in the abnormality prediction apparatus according to the first embodiment. FIG. 6 is a block diagram illustrating a configuration example of the abnormality prediction system according to the second embodiment. FIG. 7 is a block diagram illustrating a configuration example of the abnormality prediction apparatus according to the third embodiment. FIG. 8 is a diagram illustrating a computer that executes an abnormality prediction program.

  Hereinafter, embodiments of an abnormality prediction method, an abnormality prediction apparatus, an abnormality prediction system, and an abnormality prediction program according to the present application will be described in detail with reference to the drawings. Note that the present embodiment does not limit the abnormality prediction method, abnormality prediction device, abnormality prediction system, and abnormality prediction program according to the present application.

[First Embodiment]
In the following embodiments, the configuration of the abnormality prediction device 10 according to the first embodiment and the flow of processing of the abnormality prediction device 10 will be described in order, and finally the effects of the first embodiment will be described.

[Abnormality prediction device configuration]
First, the configuration of the abnormality prediction apparatus 10 will be described with reference to FIG. FIG. 1 is a block diagram illustrating a configuration example of the abnormality prediction apparatus according to the first embodiment. For example, the abnormality prediction device 10 uses the process data, which is multivariate time series data continuously collected by a sensor or the like installed in a factory or a plant, to determine whether an abnormality occurs after a predetermined time. A warning is output in advance if it is determined that an abnormality will occur.

  As illustrated in FIG. 1, the abnormality prediction device 10 includes a communication processing unit 11, an abnormality prediction unit 12, a parameter learning unit 13, and a display unit 14. Hereinafter, processing of each unit included in the abnormality prediction apparatus 10 will be described.

  The communication processing unit 11 controls communication related to various types of information exchanged with sensors installed in factories or plants. For example, the communication processing unit 11 periodically receives process data from a sensor installed in a factory or a plant.

  Further, the abnormality prediction unit 12 predicts whether or not an abnormality will occur after a predetermined time using the process data and the learning model (identification function). The abnormality prediction unit 12 includes a process data acquisition unit 12a, an analysis frame extraction unit 12b, a frame abnormality evaluation value calculation unit 12c, an abnormality prediction evaluation value calculation unit 12d, a process data buffer 12e, and a frame abnormality evaluation value buffer 12f.

  The process data acquisition unit 12a, the analysis frame extraction unit 12b, the frame abnormality evaluation value calculation unit 12c, and the abnormality prediction evaluation value calculation unit 12d are, for example, an electronic circuit such as a CPU (Central Processing Unit) or MPU (Micro Processing Unit) or an ASIC. (Application Specific Integrated Circuit) and FPGA (Field Programmable Gate Array). The process data buffer 12e and the frame abnormality evaluation value buffer 12f are storage devices such as semiconductor memory elements such as RAM (Random Access Memory) and flash memory (Flash Memory).

  The process data acquisition unit 12a acquires process data that is time-series data collected by the sensor of the monitoring target facility. Specifically, the process data acquisition unit 12a periodically acquires process data collected by sensors installed in factories, plants, and the like (for example, every minute) and stores the process data in the process data buffer 12e. Here, the data collected by the sensor is, for example, various data such as temperature, pressure, sound, vibration, and the like regarding a factory, a device in a plant, and a reactor that are monitoring target facilities.

  The analysis frame extraction unit 12b extracts process data included in an analysis frame having a predetermined time width Wa from the process data acquired by the process data acquisition unit 12a. Specifically, the analysis frame extraction unit 12b extracts and reads out process data for a frame having a time width Wa including process data at a predetermined time from the process data buffer 12e, and calculates the frame abnormality evaluation value for the read process data. Notification to the unit 12c.

  For example, in the analysis frame extraction unit 12b, when the process data acquisition unit 12a acquires process data every minute and the time width Wa is “5”, the process data acquired at the current time t, the current time The process data acquired 1 minute before, the process data acquired 2 minutes before the current time, the process data acquired 3 minutes before the current time, and the process data acquired 4 minutes before the current time Extracted from the process data buffer 12e and read.

  The process data buffer 12e stores process data acquired by the process data acquisition unit 12a. The process data buffer 12e stores at least the latest process data for a frame having a time width Wa as process data.

  The frame abnormality evaluation value calculation unit 12c receives the process data extracted by the analysis frame extraction unit 12b as an input, and uses a learning model for predicting abnormality of the monitoring target facility for a time point after a predetermined time from a predetermined time point. The frame abnormality evaluation value, which is an intermediate evaluation value representing the probability of occurrence of abnormality, is calculated. The predetermined time point may be the current time point or a preset time point such as 10 seconds before the current time point. Further, the frame abnormality evaluation value calculation unit 12c calculates a frame abnormality evaluation value using a neural network as a learning model. Note that the learning model is not limited to a neural network, and may be a machine learning algorithm such as logistic regression, support vector machine, discriminant analysis, or deep neural network.

  Specifically, the frame abnormality evaluation value calculation unit 12c receives the process data extracted from the analysis frame extraction unit 12b. Then, the frame abnormality evaluation value calculation unit 12c uses the received process data as an input, and uses a learning model for predicting abnormality of the monitoring target facility, and uses the frame abnormality evaluation value for a time point after a predetermined time from the current time t. Is calculated and stored in the frame abnormality evaluation value buffer 12f. Note that the frame abnormality evaluation value is, for example, a probability value that an abnormality of the monitoring target facility occurs, and may be a numerical value represented by “0” to “1”. In this case, for example, when the probability of occurrence of abnormality of the monitoring target facility at a certain time point is predicted to be “40%”, the frame abnormality evaluation value is “0.4”. Further, the frame abnormality evaluation value is not limited to this. For example, as a value indicating whether or not there is a possibility that abnormality of the monitoring target facility occurs, a value “0” or “1” is set. It may be a numerical value expressed by. Further, when there is time-series data such as a specific sensor value that becomes an abnormality index, the specific time-series data itself can be regarded as a frame abnormality evaluation value.

  For example, when the process data acquisition unit 12a acquires process data every minute, the time width Wa is “5”, and the time width Wb is “7”, the frame abnormality evaluation value calculation unit 12c Using the process data of the time width Wa “5” extracted by the analysis frame extraction unit 12b as an input, using a learning model for predicting the abnormality of the monitored facility, an abnormality is detected at a time point after a predetermined time from the current time. A frame abnormality evaluation value that is an intermediate evaluation value representing the probability of occurrence is calculated.

  The abnormality prediction evaluation value calculation unit 12d finally represents the probability that an abnormality will occur at a time point after a predetermined time from a predetermined time point based on the plurality of frame abnormality evaluation values calculated by the frame abnormality evaluation value calculation unit 12c. An abnormal prediction evaluation value that is a large evaluation value is calculated. For example, when a time point after a predetermined time is a time point t + N, the abnormality prediction evaluation value calculation unit 12d includes the time point t + N after the predetermined time period, and the retroactive time width Wb from the time point t + N after the predetermined time point. Is calculated as a final evaluation value representing the likelihood of occurrence of an abnormality at a time t + N after the predetermined time. Specifically, the abnormality prediction evaluation value calculation unit 12d reads the frame abnormality evaluation value from the frame abnormality evaluation value buffer 12f, and based on the frame abnormality evaluation value, the abnormality prediction at a time point t + N after a predetermined time from the current time t. An evaluation value is calculated and notified to the abnormality determination unit 14a and the chart display unit 14b.

  For example, the abnormality prediction evaluation value calculation unit 12d calculates, as an abnormality prediction evaluation value calculation method, an average value of a plurality of frame abnormality evaluation values as an abnormality prediction evaluation value. Further, the abnormality prediction evaluation value calculation unit 12d is not limited to a method of simply calculating an average value of a plurality of frame abnormality evaluation values as a method of calculating the abnormality prediction evaluation value. The calculation may be performed using various methods such as weighting.

  The frame abnormality evaluation value buffer 12f stores the frame abnormality evaluation value calculated by the frame abnormality evaluation value calculation unit 12c. The frame abnormality evaluation value buffer 12f stores at least the latest frame abnormality evaluation value for the frame having the time width Wb as the frame abnormality evaluation value.

Here, a series of abnormality prediction evaluation value calculation processing executed by the abnormality prediction unit 12 will be described with reference to FIG. FIG. 2 is a diagram illustrating an abnormality prediction evaluation value calculation process executed by the abnormality prediction unit. (A) in FIG. 2 shows a plot of the process data acquired from the sensor for each of the items x _1t to x _nt , and an analysis frame having a time width Wa every time an abnormal prediction evaluation value is calculated. Move by the moving width L.

  As illustrated in FIGS. 2A and 2B, for example, the process data acquisition unit 12 a acquires process data every minute, the time width Wa is “5”, and the time width Wb is “7”. ”. In this case, the frame abnormality evaluation value calculation unit 12c receives the process data for the time width Wa “5” as an input, and uses a learning model for predicting abnormality of the monitoring target facility for a predetermined time from the current time t. A frame abnormality evaluation value, which is an intermediate evaluation value representing the probability of occurrence of abnormality, is calculated at a later time.

  Then, as illustrated in (B) and (C) of FIG. 2, for example, the abnormality prediction evaluation value calculation unit 12d calculates the average value of each frame abnormality evaluation value at a plurality of time points “t + N” to “t + N−6”. Is calculated as an abnormality prediction evaluation value at a time “t + N” a predetermined time after the current time t. For example, the abnormality prediction evaluation value calculation unit 12d determines that each frame abnormality evaluation value of “t + N” to “t + N−4” is “0.2”, and that the frame abnormality evaluation value of “t + N−5” is “0.5”. When the frame abnormality evaluation value of “t + N−6” is “0.6”, the average value “0.3” of each frame abnormality evaluation value is set to the time “t + N” at a predetermined time after the current time t. Calculated as an abnormal prediction evaluation value. The graph shown in the lower part of FIG. 2C is obtained by plotting the abnormality prediction evaluation value, and the portion where the value is large represents the time zone in which the abnormality is predicted.

  Returning to the description of FIG. 1, the parameter learning unit 13 learns the parameters of the learning model based on the abnormality occurrence time data and the process data indicating the time when the abnormality occurs in the monitoring target facility. The parameter learning unit 13 includes a learning data storage unit 13a, a learning data extraction unit 13b, a frame abnormality evaluation value function learning unit 13c, a learning process data storage unit 13d, and a frame abnormality evaluation value function parameter storage unit 13e.

  The learning data storage unit 13a acquires process data and abnormality occurrence time data from a sensor of the monitoring target facility, an external device, or the like, and stores the acquired process data and abnormality occurrence time data in the learning process data storage unit 13d. Further, when using time series data such as a specific sensor value as a frame abnormality evaluation value, the learning data storage unit 13a stores this time series data in the learning process data storage unit 13d.

  The learning data extraction unit 13b reads the abnormality occurrence time data from the learning process data storage unit 13d, and the ratio between the data amount of the process data at the occurrence of the abnormality and the data amount of the process data at the normal time becomes a predetermined ratio. As described above, the process data is extracted from the learning process data storage unit 13d and read out, and the abnormality occurrence time data and the process data are notified to the frame abnormality evaluation value function learning unit 13c.

  That is, for example, when the data amount of the process data at the normal time is significantly larger than the data amount of the process data at the time of occurrence of the abnormality, the learning of the discriminant function cannot be performed appropriately. The process data is extracted and read from the learning process data storage unit 13d so that the ratio between the amount of process data when an abnormality occurs and the amount of process data when normal is a predetermined ratio.

  The frame abnormality evaluation value function learning unit 13c learns the parameter of the identification function for predicting the abnormality of the monitored facility based on the abnormality occurrence time data and the process data, and uses the learned identification function parameter as the frame abnormality evaluation value. Store in the function parameter storage unit 13e. For example, the frame abnormality evaluation value function learning unit 13c performs machine learning using the abnormality occurrence time data as a correct answer label. Any known machine learning method may be used. For example, a machine learning algorithm such as neural network, logistic regression, support vector machine, discriminant analysis, or deep neural network is applied. Further, when using time-series data such as a specific sensor value as a frame abnormality evaluation value, the frame abnormality evaluation value function learning unit 13c uses this time-series data as an objective variable and process data as an explanatory variable. A frame anomaly evaluation value function is learned by applying a known regression algorithm such as linear regression, support vector regression, or neural network.

  The learning process data storage unit 13d stores process data and abnormality occurrence time data acquired by the learning data storage unit 13a. The frame abnormality evaluation value function parameter storage unit 13e stores parameters of the identification function learned by the frame abnormality evaluation value function learning unit 13c.

  The display unit 14 outputs a warning sign based on the abnormality prediction evaluation value, or outputs time series data of the abnormality prediction evaluation value as a chart screen. The display unit 14 includes an abnormality determination unit 14a and a chart display unit 14b.

  The abnormality determination unit 14a determines whether or not the abnormality prediction evaluation value calculated by the abnormality prediction evaluation value calculation unit 12d is greater than or equal to a predetermined threshold. If the abnormality prediction evaluation value is greater than or equal to a predetermined threshold, Output a warning about abnormal occurrence. For example, the abnormality determination unit 14a outputs, as a warning sign, a warning message that an abnormality may occur in the monitored equipment after a certain time when the abnormality prediction evaluation value is equal to or greater than a predetermined threshold. Alternatively, a sound for notifying a warning may be output.

  The chart display unit 14b displays the time series data of the abnormality prediction evaluation value calculated by the abnormality prediction evaluation value calculation unit 12d as a chart screen. For example, when the chart display unit 14b receives a display request from the user, the chart display unit 14b displays time-series data of the abnormality prediction evaluation value as a chart screen.

  Here, it is a figure explaining the outline | summary of the abnormality prediction process performed by the abnormality prediction apparatus 10 using FIG. FIG. 3 is a diagram for explaining the outline of the abnormality prediction process executed by the abnormality prediction device.

  FIG. 3A shows that a device that collects sensors and signals for operation is attached to a reaction furnace, an apparatus, and the like in the plant, and collects data at regular intervals. FIG. 3B shows a plot of the process data collected by the sensor or the like for each item (process). The data in the portion surrounded by a thick frame is cut out, and a certain time later. It is assumed that abnormalities are estimated. 3C illustrates the abnormality prediction unit 12 that calculates the abnormality prediction evaluation value based on the process data described above. FIG. 3D shows a plot of the abnormality prediction evaluation value that is finally calculated, and a portion where the value is large represents a time zone in which the abnormality is predicted. FIG. 3E illustrates a correct answer label (abnormality occurrence time data) used for learning the discriminant function included in FIG.

  As shown in FIGS. 3A and 3B, process data, which is time-series data collected by a sensor or the like installed in an apparatus in a plant, a reaction furnace, or the like, is acquired. Then, as shown in (C) and (D) of FIG. 3, it is predicted whether or not an abnormality will occur after a predetermined time, using the process data and the identification function. The parameter learning unit 13 performs machine learning using the discriminant function abnormality occurrence time data illustrated in FIG. 3E as a correct label.

[Processing procedure of abnormality prediction device]
Next, an example of a processing procedure performed by the abnormality prediction apparatus 10 according to the first embodiment will be described with reference to FIGS. 4 and 5. FIG. 4 is a flowchart illustrating an example of a flow of abnormality prediction processing in the abnormality prediction device according to the first embodiment. FIG. 5 is a flowchart illustrating an example of a learning process flow in the abnormality prediction apparatus according to the first embodiment.

  First, the flow of the abnormality prediction process performed by the abnormality prediction device 10 will be described with reference to FIG. As illustrated in FIG. 4, when the process data acquisition unit 12a of the abnormality prediction unit 12 acquires process data (Yes in step S101), the process data acquired by the process data acquisition unit 12a has a predetermined time width Wa. Process data in a certain analysis frame is extracted (step S102).

  Then, the frame abnormality evaluation value calculation unit 12c uses the process data extracted by the analysis frame extraction unit 12b as an input, and uses a learning model for predicting abnormality of the monitoring target facility, after a predetermined time from a predetermined time point. A frame abnormality evaluation value, which is an intermediate evaluation value representing the probability of occurrence of abnormality, is calculated for the time point (step S103).

  Subsequently, the abnormality prediction evaluation value calculation unit 12d represents the probability that the abnormality will occur at a time point after a predetermined time from the current time, based on the plurality of frame abnormality evaluation values calculated by the frame abnormality evaluation value calculation unit 12c. An abnormal prediction evaluation value that is a final evaluation value is calculated (step S104).

  Thereafter, the abnormality determination unit 14a of the display unit 14 determines whether or not the abnormality prediction evaluation value calculated by the abnormality prediction evaluation value calculation unit 12d is equal to or greater than a predetermined threshold (step S105). As a result, when the abnormality prediction evaluation value is less than the predetermined threshold (No at Step S105), the abnormality determination unit 14a ends the process as it is. Moreover, the abnormality determination part 14a outputs a warning sign, when an abnormality prediction evaluation value is more than a predetermined threshold value (Yes at Step S105) (Step S106).

  Next, the flow of the learning process performed by the abnormality prediction device 10 will be described with reference to FIG. As illustrated in FIG. 5, when the learning data storage unit 13a of the parameter learning unit 13 acquires process data or abnormality occurrence time data as learning data from a sensor of a monitoring target facility or the like (Yes in step S201), the acquired learning is performed. Data is stored in the learning process data storage unit 13d (step S202). Then, the learning data extraction unit 13b determines whether or not abnormality occurrence time data has been acquired in S101 (step S203). As a result, the learning data extraction unit 13b returns to step S201 when the abnormality occurrence time data has not been acquired (No at step S203).

  Further, when the learning data extraction unit 13b acquires the abnormality occurrence time data (Yes at Step S203), the ratio between the data amount of the process data at the occurrence of the abnormality and the data amount of the process data at the normal time is a predetermined ratio. The process data is extracted from the learning process data storage unit 13d (step S204).

  Subsequently, the frame abnormality evaluation value function learning unit 13c learns the parameters of the identification function for predicting the abnormality of the monitoring target facility based on the process data and abnormality occurrence time data extracted by the learning data extraction unit 13b. (Step S205). Then, the frame abnormality evaluation value function learning unit 13c stores the learned identification function parameter in the frame abnormality evaluation value function parameter storage unit 13e (step S206).

[Effect of the first embodiment]
The abnormality prediction device 10 according to the first embodiment acquires process data that is time-series data collected by a sensor of a monitoring target facility, and an analysis frame having a predetermined time width Wa from the acquired process data. Extract process data contained in. Then, the abnormality prediction device 10 uses the extracted process data as an input, and uses a learning model for predicting abnormality of the monitoring target facility, and is likely to cause abnormality at a time point after a predetermined time from a predetermined time point. A frame abnormality evaluation value that is an intermediate evaluation value that represents Then, the abnormality prediction device 10 is based on the calculated plurality of frame abnormality evaluation values, and the abnormality prediction evaluation value that is a final evaluation value that represents the probability that an abnormality will occur at a time point after a predetermined time from a predetermined time point. Is calculated. For this reason, the abnormality prediction apparatus 10 can predict whether an abnormality will occur in the future.

  For example, the abnormality prediction apparatus 10 calculates an abnormality prediction evaluation value that is a final evaluation value based on a plurality of frame abnormality evaluation values, and therefore removes noise from frame abnormality evaluation values that have a large variation at successive times. In order to obtain a stable and stable output, it is possible to perform smoothing and the like, calculate an accurate abnormality prediction evaluation value, and predict whether or not an abnormality will occur after a preset period of time. .

  The abnormality prediction device 10 according to the first embodiment determines whether or not the abnormality prediction evaluation value is equal to or greater than a predetermined threshold, and if the abnormality prediction evaluation value is equal to or larger than the predetermined threshold, a warning regarding the occurrence of abnormality. Is output. For this reason, the abnormality prediction device 10 can output a warning by predicting that an abnormality will occur after a certain time, and can take measures such as avoiding the abnormality in advance.

  The abnormality prediction device 10 according to the first embodiment displays time series data of abnormality prediction evaluation values as a chart screen. For this reason, the abnormality prediction device 10 can confirm the transition of the abnormality prediction evaluation value, and can cope with the abnormality in advance.

  The abnormality prediction device 10 according to the first embodiment calculates a frame abnormality evaluation value using a neural network as a learning model. For this reason, the abnormality prediction device 10 can appropriately calculate the frame abnormality evaluation value by using a known neural network.

  The abnormality prediction device 10 according to the first embodiment learns parameters of a learning model based on abnormality occurrence time data and process data indicating a time when an abnormality has occurred in a monitoring target facility. For this reason, the abnormality prediction device 10 can learn the identification function using the process data collected in the past and the information on the time of occurrence of the abnormality, and can predict the abnormality corresponding to the temporal change of the monitored equipment. Become.

[Second Embodiment]
In the first embodiment described above, the abnormality prediction apparatus 10 includes the abnormality prediction unit 12 and the parameter learning unit 13, and performs a process of calculating an abnormality prediction evaluation value and a process of learning a parameter of an identification function. Although described, the present invention is not limited to this. For example, the process of learning the parameters of the identification function may be performed in a remote environment via a communication line.

  Therefore, in the following, the abnormality prediction device 10A according to the second embodiment performs processing for calculating the abnormality prediction evaluation value, and the parameter learning device 20 in the remote environment via the communication line learns the parameters of the identification function. The case of performing will be described. In addition, description is abbreviate | omitted about the structure and process similar to the abnormality prediction apparatus 10 which concerns on 1st Embodiment.

  First, the abnormality prediction system 100 according to the second embodiment will be described with reference to FIG. FIG. 6 is a block diagram illustrating a configuration example of the abnormality prediction system according to the second embodiment. As illustrated in FIG. 6, the abnormality prediction system 100 includes an abnormality prediction device 10 </ b> A in a local environment and a parameter learning device 20 in a remote environment. In the abnormality prediction system 100, the abnormality prediction device 10A in the local environment and the parameter learning device 20 in the remote environment are connected via a network 30.

  As shown in FIG. 6, the abnormality prediction apparatus 10A according to the second embodiment has a parameter learning unit 13 as compared with the abnormality prediction apparatus 10 according to the first embodiment shown in FIG. However, the communication unit 15 is different.

  The communication unit 15 transmits process data and abnormality occurrence time data to the parameter learning device 20 and receives parameters of the learning model for predicting abnormality of the monitoring target facility from the parameter learning device 20. The communication unit 15 includes a process data transmission / reception unit 15a, a process data storage unit 15b, a parameter reception unit 15c, and a local frame abnormality evaluation value function parameter storage unit 15d.

  When the process data transmission / reception unit 15a receives process data from a sensor or the like of the monitoring target facility, the process data transmission / reception unit 15a stores the received process data in the process data storage unit 15b and transmits the process data to the parameter learning device 20. Further, the process data transmission / reception unit 15a transmits the abnormality occurrence time data to the parameter learning device 20 when receiving the abnormality occurrence time data from an external device or the like. The process data storage unit 15b stores the process data received by the process data transmission / reception unit 15a.

  The parameter receiving unit 15c receives the parameter of the discriminant function learned by the parameter learning device 20, and stores the received parameter of the discriminant function in the local frame abnormality evaluation value function parameter storage unit 15d. The local frame abnormality evaluation value function parameter storage unit 15d stores the parameter of the identification function received by the parameter reception unit 15c.

  The frame abnormality evaluation value calculation unit 12c reads the parameter of the identification function from the local frame abnormality evaluation value function parameter storage unit 15d. Then, similarly to the first embodiment, the frame abnormality evaluation value calculation unit 12c receives the process data extracted by the analysis frame extraction unit 12b as an input, and uses a discriminant function to determine a time point after a predetermined time. A frame abnormality evaluation value, which is an intermediate evaluation value representing the probability of occurrence of abnormality, is calculated.

  The parameter learning device 20 does not have the learning data storage unit 13a as compared with the parameter learning unit 13 according to the first embodiment shown in FIG. 1, but the learning data reception unit 13f and the parameter transmission unit 13g It has different points. The learning data receiving unit 13f receives process data and abnormality occurrence time data from the abnormality prediction device 10A. When the learning data receiving unit 13f receives the process data and the abnormality occurrence time data, the learning data receiving unit 13f stores the received process data and abnormality occurrence time data in the learning process data storage unit 13d.

  The parameter transmission unit 13g transmits the parameter of the discrimination function learned by the frame abnormality evaluation value function learning unit 13c to the abnormality prediction device 10A. Specifically, the parameter transmission unit 13g reads the identification function parameter stored in the frame abnormality evaluation value function parameter storage unit 13e, and transmits the read identification function parameter to the abnormality prediction device 10A.

[Effects of Second Embodiment]
In the abnormality prediction system 100 according to the second embodiment, the abnormality prediction device 10A performs a process of calculating an abnormality prediction evaluation value, and the parameter learning device 20 in the remote environment via the communication line learns the parameters of the discrimination function. Process. Thereby, for example, it is possible to learn the parameters of the discrimination function with a high-performance computer in a remote place connected by a communication line, and to perform highly accurate learning processing. Further, for example, when there are few opportunities for occurrence of an abnormality and the frequency of performing the learning process is low, it is possible to cause the apparatus outside the remote environment to perform the learning process instead of the abnormality prediction apparatus 10A in the local environment. .

[Third Embodiment]
In the first embodiment and the second embodiment described above, when the learning model parameter is processed and the learning model parameter is updated, an abnormality is detected using the learning model to which the latest learning parameter is applied. Although the case where the predicted evaluation value is calculated has been described, the present invention is not limited to this. For example, the abnormality prediction evaluation value may be calculated using a learning model that has been learned in advance without performing the process of learning the parameters of the learning model.

  Therefore, hereinafter, a case will be described in which the abnormality prediction device 10B according to the third embodiment calculates an abnormality prediction evaluation value using a learning model that has been learned in advance. In addition, description is abbreviate | omitted about the structure and process similar to the abnormality prediction apparatus 10 which concerns on 1st Embodiment.

  First, a configuration example of the abnormality prediction apparatus 10B according to the third embodiment will be described with reference to FIG. FIG. 7 is a block diagram illustrating a configuration example of the abnormality prediction apparatus according to the third embodiment. As illustrated in FIG. 7, the abnormality prediction device 10 </ b> B according to the third embodiment has no parameter learning unit 13 as compared with the abnormality prediction device 10 according to the first embodiment illustrated in FIG. 1. Different.

  The frame abnormality evaluation value calculation unit 12c receives the process data extracted by the analysis frame extraction unit 12b, and an abnormality occurs at a time point after a predetermined time from a predetermined time point using a learning model learned in advance. A frame abnormality evaluation value that is an intermediate evaluation value representing the certainty is calculated.

[Effect of the third embodiment]
In the abnormality prediction device 10B according to the third embodiment, the abnormality prediction evaluation value is calculated using a learning model that has been learned in advance, so that it is predicted whether an abnormality will occur in the future while suppressing the processing load. It is possible.

(System configuration etc.)
Further, each component of each illustrated apparatus is functionally conceptual, and does not necessarily need to be physically configured as illustrated. In other words, the specific form of distribution / integration of each device is not limited to that shown in the figure, and all or a part thereof may be functionally or physically distributed or arbitrarily distributed in arbitrary units according to various loads or usage conditions. Can be integrated and configured. Further, all or any part of each processing function performed in each device may be realized by a CPU and a program analyzed and executed by the CPU, or may be realized as hardware by wired logic.

  In addition, among the processes described in this embodiment, all or part of the processes described as being automatically performed can be manually performed, or the processes described as being manually performed All or a part of the above can be automatically performed by a known method. In addition, the processing procedure, control procedure, specific name, and information including various data and parameters shown in the above-described document and drawings can be arbitrarily changed unless otherwise specified.

(program)
It is also possible to create a program in which the processing executed by the abnormality prediction apparatus described in the above embodiment is described in a language that can be executed by a computer. For example, an abnormality prediction program in which processing executed by the abnormality prediction apparatus 10 according to the embodiment is described in a language that can be executed by a computer can be created. In this case, when the computer executes the abnormality prediction program, the same effect as in the above embodiment can be obtained. Further, even if the abnormality prediction program is recorded on a computer-readable recording medium, and the abnormality prediction program recorded on the recording medium is read by the computer and executed, the same processing as in the above embodiment can be realized. Good.

  FIG. 8 is a diagram illustrating a computer that executes an abnormality prediction program. As illustrated in FIG. 8, the computer 1000 includes, for example, a memory 1010, a CPU 1020, a hard disk drive interface 1030, a disk drive interface 1040, a serial port interface 1050, a video adapter 1060, and a network interface 1070. These units are connected by a bus 1080.

  The memory 1010 includes a ROM (Read Only Memory) 1011 and a RAM 1012 as illustrated in FIG. The ROM 1011 stores a boot program such as BIOS (Basic Input Output System). The hard disk drive interface 1030 is connected to the hard disk drive 1090 as illustrated in FIG. The disk drive interface 1040 is connected to the disk drive 1100 as illustrated in FIG. For example, a removable storage medium such as a magnetic disk or an optical disk is inserted into the disk drive 1100. The serial port interface 1050 is connected to, for example, a mouse 1110 and a keyboard 1120 as illustrated in FIG. The video adapter 1060 is connected to a display 1130, for example, as illustrated in FIG.

  Here, as illustrated in FIG. 8, the hard disk drive 1090 stores, for example, an OS 1091, an application program 1092, a program module 1093, and program data 1094. That is, the abnormality prediction program is stored in, for example, the hard disk drive 1090 as a program module in which a command to be executed by the computer 1000 is described.

  In addition, various data described in the above embodiment is stored as program data in, for example, the memory 1010 or the hard disk drive 1090. Then, the CPU 1020 reads the program module 1093 and the program data 1094 stored in the memory 1010 and the hard disk drive 1090 to the RAM 1012 as necessary, and executes various processing procedures.

  The program module 1093 and the program data 1094 related to the abnormality prediction program are not limited to being stored in the hard disk drive 1090, but are stored in, for example, a removable storage medium and read out by the CPU 1020 via the disk drive or the like. Also good. Alternatively, the program module 1093 and the program data 1094 related to the abnormality prediction program are stored in another computer connected via a network (LAN (Local Area Network), WAN (Wide Area Network), etc.), and the network interface 1070 is stored. Via the CPU 1020.

  The above embodiments and modifications thereof are included in the invention disclosed in the claims and equivalents thereof as well as included in the technology disclosed in the present application.

10, 10A, 10B Abnormality prediction device 11 Communication processing unit 12 Abnormality prediction unit 12a Process data acquisition unit 12b Analysis frame extraction unit 12c Frame abnormality evaluation value calculation unit 12d Abnormality prediction evaluation value calculation unit 12e Process data buffer 12f Frame abnormality evaluation value buffer 13 parameter learning unit 13a learning data storage unit 13b learning data extraction unit 13c frame abnormality evaluation value function learning unit 13d learning process data storage unit 13e frame abnormality evaluation value function parameter storage unit 13f learning data reception unit 13g parameter transmission unit 14 display unit 14a abnormality determination unit 14b chart display unit 15 communication unit 15a process data transmission / reception unit 15b process data storage unit 15c parameter reception unit 15d local frame abnormality evaluation value function parameter storage unit 20 Parameter learning apparatus 30 network 100 abnormality prediction system

Claims (8)

An abnormality prediction method executed by an abnormality prediction device,
An extraction step of extracting process data included in a first time width that is a predetermined time width from process data that is time-series data collected by a sensor of a monitoring target facility;
An intermediate representing the probability of occurrence of an abnormality at a time point after a predetermined time from a predetermined time point using a learning model for predicting the abnormality of the monitored facility using the process data extracted by the extraction step as an input A frame abnormality evaluation value calculating step for calculating a frame abnormality evaluation value which is a typical evaluation value;
Based on a plurality of frame abnormality evaluation values calculated by the frame abnormality evaluation value calculating step, an abnormality prediction evaluation that is a final evaluation value representing the probability that an abnormality will occur at a time point after a predetermined time from the predetermined time point. An abnormality prediction method comprising: an abnormality prediction evaluation value calculation step for calculating a value.   It is determined whether or not the abnormality prediction evaluation value calculated by the abnormality prediction evaluation value calculation step is equal to or greater than a predetermined threshold, and when the abnormality prediction evaluation value is equal to or larger than a predetermined threshold, a warning about occurrence of abnormality The abnormality prediction method according to claim 1, further comprising an abnormality determination step of outputting.   The abnormality prediction method according to claim 1, further comprising a chart display step of displaying time series data of the abnormality prediction evaluation value calculated by the abnormality prediction evaluation value calculation step as a chart screen.   The abnormality prediction method according to claim 1, wherein the frame abnormality evaluation value calculation step calculates the frame abnormality evaluation value using a neural network as the learning model.   2. The learning method according to claim 1, further comprising a learning step of learning the learning model parameters based on the abnormality occurrence time data indicating the time when the abnormality occurs in the monitoring target facility and the process data. Anomaly prediction method. Extraction means for extracting process data included in a first time width that is a predetermined time width from process data that is time-series data collected by a sensor of a monitoring target facility;
An intermediate representing the probability of occurrence of an abnormality at a time point after a predetermined time from a predetermined time point using a learning model for predicting the abnormality of the monitored facility using the process data extracted by the extraction means as an input Frame abnormality evaluation value calculating means for calculating a frame abnormality evaluation value which is a typical evaluation value;
Based on a plurality of frame abnormality evaluation values calculated by the frame abnormality evaluation value calculation means, an abnormality prediction evaluation that is a final evaluation value representing the probability that an abnormality will occur at a time after a predetermined time from the predetermined time An abnormality prediction apparatus comprising: an abnormality prediction evaluation value calculation means for calculating a value. An anomaly prediction system in which an anomaly prediction device and a parameter learning device are connected via a network,
The abnormality prediction device is
Process data that is time-series data collected by sensors of the monitoring target equipment and abnormality occurrence time data indicating the time when the abnormality occurred in the monitoring target equipment is transmitted to the parameter learning device, and the abnormality of the monitoring target equipment Communication means for receiving parameters of a learning model for predicting from the parameter learning device;
Extraction means for extracting process data included in a first time width which is a predetermined time width from the process data;
A frame abnormality evaluation value that is an intermediate evaluation value representing the probability of occurrence of an abnormality at a time point after a predetermined time from a predetermined time point using the learning model as an input with the process data extracted by the extraction unit A frame abnormality evaluation value calculating means for calculating
Based on a plurality of frame abnormality evaluation values calculated by the frame abnormality evaluation value calculation means, an abnormality prediction evaluation that is a final evaluation value representing the probability that an abnormality will occur at a time after a predetermined time from the predetermined time An abnormality prediction evaluation value calculation means for calculating a value,
The parameter learning device is
Learning data receiving means for receiving the process data and the abnormality occurrence time data from the abnormality prediction device;
Learning means for learning parameters of the learning model based on the process data and the abnormality occurrence time data received by the learning data receiving means;
An abnormality prediction system comprising: parameter transmission means for transmitting parameters of the learning model learned by the learning means to the abnormality prediction device. An extraction step for extracting process data included in a first time width that is a predetermined time width from process data that is time-series data collected by a sensor of a monitoring target facility;
An intermediate representing the probability that an abnormality will occur at a time after a predetermined time from a predetermined time using the learning model for predicting the abnormality of the monitored equipment, using the process data extracted by the extraction step as an input A frame abnormality evaluation value calculating step for calculating a frame abnormality evaluation value which is a typical evaluation value;
Based on the plurality of frame abnormality evaluation values calculated by the frame abnormality evaluation value calculating step, an abnormality prediction evaluation that is a final evaluation value representing the probability that an abnormality will occur at a time after a predetermined time from the predetermined time An abnormality prediction program that causes a computer to execute an abnormality prediction evaluation value calculation step for calculating a value.

Images