JP7042082B2 - Abnormality sign diagnosis device, abnormality sign diagnosis method and abnormality sign diagnosis program - Google Patents

Description

The present invention relates to an abnormality sign diagnosis device, an abnormality sign diagnosis method, and an abnormality sign diagnosis program.

If a device deteriorates and cannot operate normally, the device will be repaired or updated. There is no problem if repairs and updates can be carried out systematically. However, if the equipment installed in a large-scale production line suddenly stops, its repair / renewal requires unexpected labor and cost, which has a great impact on corporate management. Therefore, it is important to detect the sign and preventively repair / update the device before it falls into a full-scale abnormality.

Prior to the diagnosis of the plant, the diagnostic apparatus of Patent Document 1 constructs a normal model using the state quantity acquired from the plant at the time when the plant is known to be normal. The diagnostic device determines whether or not the state quantity acquired from the plant at the time of diagnosis is classified as a normal model. The diagnostic device indicates that the plant is normal when the state quantity acquired from the plant at the time of diagnosis is classified as a normal model. If the state quantity acquired from the plant at the time of diagnosis is not classified as a normal model, the diagnostic device indicates that the plant is in an unknown state that has never been experienced in the past.

International Publication No. 2012/073289

In an actual plant, the same process is often performed in parallel. For example, a plurality of processes (several tens to several hundreds in the case of many) in which raw materials A and B are mixed in a reactor to produce product C are arranged in parallel, and some or all of them are in operation according to a production plan. do. In such a case, when an abnormality occurs in one of the parallel processes, the same type of abnormality often occurs in the other processes. As a plant maintainer, I would like to know at once whether or not an abnormality has occurred in the processes in parallel, and if so, what kind of abnormality has occurred.

However, in the diagnostic apparatus of Patent Document 1, even when a plurality of processes exist in the plant, which process and which process are in parallel, and which process and which process are in series. I do not particularly recognize the relationship. As a result, the diagnostic apparatus individually outputs diagnostic results of a plurality of parallel processes. After all, when a maintenance person tries to diagnose each of a plurality of parallel processes, the number of diagnoses is multiple, and the diagnosis results of each cannot be listed, and it is also possible to know the commonality between the diagnosis results. Can not.
Therefore, an object of the present invention is to efficiently perform diagnosis of a plurality of parallel processes.

The abnormality sign diagnosis device of the present invention acquires the measured values of a plurality of parallel processes in which the same type of device performs the same type of processing in time series for each of the parallel processes, and among the measured values acquired from each of the parallel processes. It is considered that the pre-process unit that associates a plurality of different virtual times included in a predetermined time width with respect to the ones having the same actual time actually acquired, and the measured values acquired at the virtual time. Other means, characterized in that they are provided with a classification unit, will be described in the form for carrying out the invention.

According to the present invention, it is possible to efficiently diagnose a plurality of parallel processes.

It is a figure explaining the structure of the abnormality sign diagnosis device. It is a figure explaining the structure of a plant. It is a figure explaining the classification of the measured value. It is a figure explaining the real time and the virtual time. It is a figure explaining the measured value information. It is a figure explaining the measured value information after sorting. It is a figure explaining the effect of the virtual time and the introduction of a virtual process. It is a figure explaining the effect of the virtual time and the introduction of a virtual process. It is a figure explaining the clustering using the adaptive resonance theory. It is a figure explaining the degree of abnormality and the degree of contribution. It is a flowchart of a processing procedure. It is a figure explaining the abnormality degree / contribution degree information. It is a figure explaining the diagnosis result information.

Hereinafter, a mode for carrying out the present invention (referred to as “the present embodiment”) will be described in detail with reference to figures and the like. This embodiment is an example of diagnosing a plant that produces a chemical product. However, the present invention is also applicable to other objects.

(Configuration of abnormality sign diagnosis device)
The configuration of the abnormality sign diagnosis device 1 will be described with reference to FIG. The abnormality sign diagnosis device 1 is a general computer, and includes a central control device 11, an input device 12, an output device 13, a main storage device 14, an auxiliary storage device 15, and a communication device 16. These are interconnected by a bus. The auxiliary storage device 15 stores the measurement value information 31, the abnormality degree / contribution degree information 32, and the diagnosis result information 33.

The pre-processing unit 21, the classification unit 22, and the post-processing unit 23 in the main storage device 14 are programs. In the following, when the operation subject is described as "○○ part is", it is that the central control unit 11 reads out the information used for the processing in the "○○ part" from the auxiliary storage device 15 and stores it in the main storage device 14. After that, it means to execute the process described later. The abnormality sign diagnosis device 1 is usually arranged in a plant monitoring center (not shown) or the like.

The plant 2 of the present embodiment is a set of equipment for producing chemical products, and has a plurality of parallel steps (details will be described later). The measurement / control device 3 acquires measured values such as flow rate, temperature, and pressure from sensors and the like arranged in various parts of the plant 2, and directly predicts the measured values via the network 4 or not via the network 4. It is transmitted to the diagnostic device 1. The measurement / control device 3 controls the opening degree of the valve arranged in the piping of the plant 2 so that the measured values such as the flow rate, the temperature, and the pressure become those target values.

(plant)
The configuration of the plant 2 will be described with reference to FIG. As a whole, the plant 2 produces the product C by reacting the raw material A and the raw material B. More specifically, process PR1, process PR2 and process PR3 share this production in parallel. A tank 41 for storing the raw material A and a tank 42 for storing the raw material B exist upstream of the process PR1, the process PR2, and the process PR3, and a tank 43 for storing the product C exists downstream.

Pay attention to the process PR1. The raw material A is sent from the tank 41 to the reactor 46 via the pump 44A and the valve 45A. The raw material B is sent from the tank 42 to the reactor 46 via the pump 44B and the valve 45B. The reaction apparatus 46 reacts the raw material A and the raw material B to produce the product C. Product C is sent to the tank 43. The reaction rate of the reaction device 46 varies depending on the temperature of the reaction device 46. The reaction device 46 has a cooling device (not shown), and the flow rate of the cooling water is controlled by the opening degree of the valve 47.

A flow meter F1A is arranged in the pipe between the valve 45A and the reaction device 46. The flow meter F1A measures the flow rate of the raw material A, and the measurement / control device 3 (FIG. 1) controls the opening degree of the valve 45A based on the measured value. A flow meter F1B is arranged in the pipe between the valve 45B and the reaction device 46. The flow meter F1B measures the flow rate of the raw material B, and the measurement / control device 3 controls the opening degree of the valve 45B based on the measured value. The pressure gauge P1 measures the pressure of the reaction device 46. A thermometer T1 is arranged at the outlet of the reactor 46. The thermometer T1 measures the temperature at the outlet of the reaction device 46, and the measurement / control device 3 controls the opening degree of the valve 47 based on the measured value. Then, the flow rate of the cooling water flowing into the reaction device 46 changes.
The same applies to the steps PR2 and PR3.

(Multiple parallel processes of the same)
As is clear from the above, the parallel steps in the present embodiment are a plurality of steps having the following characteristics, and are hereinafter also referred to as "parallel steps".
-Each of the steps performs the same kind of processing (for example, processing in which raw material A and raw material B are reacted to produce product C).
-Each of the steps has the same type of equipment (eg, pumps, valves and reactors).
-Each process is arranged in parallel with other processes.
• In each of the steps, the sensor acquires the same type of state quantity (eg, flow rate, pressure and temperature).

(Classification of measured values)
The classification of measured values will be described with reference to FIG. The abnormality sign diagnostic device 1 of the present embodiment classifies the measured values into a plurality of categories by using the Adaptive Resonance Theory, which is one of the clustering techniques. The classification using the adaptive resonance theory is also described in Patent Document 1. The measured value is generally a time series vector. Each component of the vector is a measured value such as flow rate, pressure, and temperature. The abnormality sign diagnosis device 1 draws points indicating these vectors in a multidimensional space, and classifies those whose positions are close to each other into the same category.

The horizontal axis of FIG. 3 is time in both the upper figure 51 and the lower figure 52. The vertical axis of FIG. 3 is a measured value in the upper figure 51, and is a category in the lower figure 52. The time on the horizontal axis is divided into a normal period in which the plant condition is known to be normal, and a diagnostic period in which it is not known whether the plant condition is normal or not. The normal period is also the period for collecting a sample of the measured values that serves as a reference for comparing the measured values during the diagnosis period. In that sense, the normal period is also called the "learning" period. In FIG. 51 above, a solid line graph FA showing a change in the flow rate of the raw material A in a certain process and a broken line graph FB showing the change in the flow rate of the raw material B in the process are drawn. The pressure of the reactor and the outlet temperature are also measured at the same time. However, for the sake of brevity, these are omitted here.

The graph FA and the graph FB move up and down between the warning upper limit value and the warning lower limit value common to both. When the range from the lower limit of the warning to the upper limit of the warning is divided into "large" and "small" with the intermediate value as the boundary line, the combination of four types of "large" and "small" is assumed as shown below. Can be done.
(FA, FB) = (Large, Small), (Small, Small), (Small, Large), (Large, Large)

At six time points of period 53 of the normal period, the combination was "(FA, FB) = (large, small)". Therefore, the abnormality sign diagnosis device 1 classifies this combination into "category 1", and six "◆" which are the classification results are drawn in FIG. 52 below. At four time points of period 54 of the normal period, the combination was "(FA, FB) = (small, small)". Therefore, the abnormality sign diagnosis device 1 classifies this combination into "category 2", and four "◆" which are the classification results are drawn in FIG. 52 below. At two time points of period 55 of the normal period, the combination was "(FA, FB) = (small, large)". Therefore, the abnormality sign diagnosis device 1 classifies this combination into "category 3", and two "◆" which are the classification results are drawn in FIG. 52 below.

In the normal period, the abnormality sign diagnosis device 1 classifies the measured values into one of category 1, category 2, and category 3. Differences in measured values between these categories are due to differences in operating conditions, operating environment, and the like.

At three time points of period 56 of the diagnosis period, the combination was "(FA, FB) = (small, small)". Therefore, the abnormality sign diagnosis device 1 classifies this combination into the known (learned) "category 2", and three "◆" which are the classification results are drawn in FIG. 52 below. At three time points of period 57 of the diagnosis period, the combination was "(FA, FB) = (large, large)". This combination has not appeared in the normal period. Therefore, the abnormality sign diagnosis device 1 classifies this combination into a new "category 4", and three "■" which are the classification results are drawn in FIG. 52 below. The abnormality sign diagnosis device 1 classifies the combination of measured values based on the magnitude relationship between the two types of measured values as described above, and also classifies the combination of measured values based on the difference between the two types of measured values. May be good.

In existing technology, such classification is performed for each of the parallel processes. For convenience of explanation, FIGS. 4 to 7 are skipped and the process proceeds to FIG. 8 temporarily. The upper left figure 81, the middle left figure 82, and the lower left figure 83 of FIG. 8 illustrate the relationship between the two types of measured values in the upper figure 51 of FIG. 3 and the four categories in the lower figure 52. Of these, the upper left FIG. 81 corresponds to the process PR1 of FIG. 2, the middle left FIG. 82 corresponds to the process PR2 of FIG. 2, and the lower left FIG. 83 corresponds to the process PR3 of FIG.

The horizontal axis of the coordinate plane of FIG. 81 in the upper left of FIG. 8 is the flow rate FA of the raw material A, and the vertical axis is the flow rate FB of the raw material B. Four spheres (clusters) indicating category 1, category 2, category 3, and category 4 are drawn on the coordinate plane. The "sphere" in the present embodiment also includes a two-dimensional circle. Of these, each of the spheres showing category 1, category 2 and category 3 contains many points (not shown) showing the measured value "(FA, FB)" in the normal period. The sphere indicating category 4 includes a point “⊚” indicating a measured value during the diagnosis period. The user (plant maintainer) of the abnormality sign diagnosis device 1 sees FIG. 81 on the upper left and knows that an unknown state has occurred in the process PR1. The user naturally wants to know the states of the process PR2 and the process PR3. However, with the existing technology, it is necessary to switch the screen for each parallel process for that purpose. It would be convenient if there was a right figure 84 that could list all the processes. The explanation returns to FIG.

(Real time and virtual time)
The real time and the virtual time will be described with reference to FIG. The actual time is the actual time when the measured value is measured by a sensor or the like. In the present embodiment, four types of measured values are simultaneously measured in three parallel steps at a certain real time. The virtual time is a virtual time in information processing used by the abnormality sign diagnosis device 1 of the present embodiment for the purpose of simultaneously diagnosing parallel processes. The real time is represented by "t + 1 digit" such as "t1", and the virtual time is represented by "t + 2 digits" such as "t11".

Now, the flow rate of the raw material A of the process PR1 actually acquired at a certain time point t1 in the diagnosis period is expressed as "F1A (t1)". Here, "t1" is the real time. Similarly, the flow rate of the raw material B of the step PR1 actually acquired at the time point t1 is expressed as “F1B (t1)”. The pressure of the reactor of the step PR1 actually acquired at the time point t1 is expressed as "P1 (t1)". The temperature at the outlet of the reactor of the step PR1 actually acquired at the time point t1 is expressed as "T1 (t1)".

The vector "(F1A (t1), F1B (t1), P1 (t1), T1 (t1))" having these as components is called a measured value vector. Similarly, the measured value vector "(F2A (t1), F2B (t1), P2 (t1), T2 (t1))" of the process PR2 at the time point t1 is defined. Further, the measurement value vector "(F3A (t1), F3B (t1), P3 (t1), T3 (t1))" of the process PR3 at the time point t1 is also defined. The measured value vector is similarly defined for the real time other than t1. The horizontal axis of FIG. 61 above FIG. 4 is the actual time. Each of the broken line rectangles in FIG. 61 above FIG. 4 is defined as a measured value vector.

The abnormality sign diagnosis device 1 of the present embodiment replaces the real time “t1” included in the measured value vector 63 of the step PR1 in FIG. 61 above with the virtual time “t11”. Similarly, the abnormality sign diagnosis device 1 replaces the real time “t1” included in the measured value vector 64 of the process PR2 with the virtual time “t12”, and replaces the real time “t1” included in the measured value vector 65 of the process PR3 with the virtual time “t12”. Replace with virtual time "t13". The abnormality sign diagnosis device 1 performs the same processing for the real times “t2” and “t3”.

After that, the abnormality sign diagnosis device 1 arranges the measured value vectors in which the real time is replaced with the virtual time in time series in FIG. 62 below of FIG. The horizontal axis of FIG. 62 below of FIG. 4 is the virtual time. At this time, it is assumed that "t1 = t11 <t12 <t13 <t2 = t21 <t22 <t23 <t3 = t31 <t32 <t33" is established. In FIG. 62 below of FIG. 4, the lengths between t11, t12, t13, t21, t22, t23, t31, t32 and t33 are arbitrary and need not be equal to each other, but may be evenly spaced. Desirable (details below).

(Measured value information)
The measured value information 31 will be described with reference to FIG. In the measured value information 31, the measured value ID is associated with the process ID stored in the process ID column 101, the measured value type column 103 is associated with the measured value type, and the measured value type column 103 is associated with the measured value type. The actual time is stored in, and the measured value is stored in the measured value column 105.
The process ID in the process ID column 101 is an identifier that uniquely identifies a parallel process.
The measured value ID in the measured value ID column 102 is an identifier that uniquely identifies the measured value. The measured value ID of the present embodiment is, for example, "F1A (t1)", so that it is immediately possible to know which kind of measured value of which process at which real time.

The type of the measured value in the measured value type column 103 is one of "raw material A flow rate", "raw material B flow rate", "reactor pressure", and "reactor outlet temperature". As a matter of course, the type of measured value depends on the contents of the plant. The ones listed here are just examples.
The real time in the real time column 104 is the above-mentioned real time. The flow rate is defined by volume or mass per unit time. That is, a certain time width (for example, 5 seconds) is required to measure the flow rate. The actual time here indicates the time of the starting point of such a time width.
The measured value in the measured value column 105 is the measured value itself measured by the sensor or the like. “#” Abbreviates different values. It is preferable that the abnormality sign diagnosis device 1 normalizes the measured values measured by the same type of sensors across a plurality of parallel processes, for example, in the range of 0 to 1. This is because the range of values that can be taken in each parallel process may differ even if the measured values are measured at the same position in the same type of device.

Looking at FIG. 5, for example, the following can be seen.
-There are three parallel processes, PR1, PR2, and PR3.
・ Four types of measured values in each process, such as F1A (t1), F1B (t1), P1 (t1) and T1 (t1), every minute after 10:00:00 on October 10, 2017. Are being measured at the same time.

(Measured value information after sorting)
The sorted measurement value information 31b will be described with reference to FIG. In the sorted measurement value information 31b, the measurement value ID is associated with the process ID stored in the process ID column 111, and the measurement value type is virtual in the measurement value type column 113. The virtual time is stored in the time column 114, the measured value is stored in the measured value column 115, and the original measured value ID is stored in the original measured value ID column 116.

The process ID in the process ID column 111 is the same as the process ID in FIG. However, here, the process IDs of all the records (rows) are "VPR1" indicating "virtual process". The term "virtual process" indicates that the process PR1, the process PR2, and the process PR3, in which the same type of device performs the same type of processing, are regarded as one process as a whole.

The measured value ID in the measured value ID column 112 is the same as the measured value ID in FIG. However, the measured value ID here, for example, "F1A (t11)", makes it possible to immediately know which kind of measured value of which process at which virtual time.
The type of the measured value in the measured value type column 113 is any one of "raw material A flow rate", "raw material B flow rate", "reactor pressure", and "reactor outlet temperature".

The virtual time in the virtual time field 114 is the virtual time described above. As described above, the virtual time is a virtual time for causing the abnormality sign diagnosis device 1 to process information, and the abnormality sign diagnosis device 1 considers that the measured value is acquired at the virtual time.
The measured value in the measured value column 115 is the same as the measured value in FIG.
The original measurement value ID in the original measurement value ID column 116 is the measurement value ID (measurement value ID before the real time is replaced with the virtual time) of the record of FIG. 5 corresponding to the record.

(Recognition of parallel process)
The abnormality sign diagnosis device 1 reads the types of measured values (raw material A flow rate, raw material B flow rate, reaction device pressure, and reaction device outlet temperature) in the measured value information 31 (FIG. 5). Then, as a result of recognizing that these are the same across the processes PR1, PR2 and PR3, the abnormality sign diagnosis device 1 recognizes that these are parallel processes, and assigns the process ID "VPR1" of the virtual process to these. Give. In addition, the abnormality sign diagnosis device 1 may recognize the parallel process by accepting, for example, a user inputting a plurality of measured value IDs via the input device 12. In FIG. 6, a double line is drawn at the boundary of the real time (t1, t2, and t3) included in the original measured value ID. This is solely for the sake of easy understanding of the explanation, and is irrelevant to the processing content of the abnormality sign diagnosis device 1.

Comparing FIGS. 5 and 6, for example, the following can be seen.
The real time "t1" for the process PR1 is replaced with the virtual time "t11". “T1” is “October 10, 2017 10:00:00”. “T11” is also “October 10, 2017 10:00:00”. The abnormality signing device 1 selects any one of the parallel processes, and for the process, the virtual time is made to match the real time.

The real time "t1" for the process PR2 is replaced with the virtual time "t12". "T1" is "October 10, 2017 10:00:20", while "t12" is "October 10, 2017 10:00:20". The virtual time is 20 seconds behind the real time. This delay will be described later.

The real time "t1" for the process PR3 is replaced with the virtual time "t13". "T1" is "October 10, 2017 10:00:00", while "t13" is "October 10, 2017 10:00:40". The virtual time is 40 seconds behind the real time. This delay will also be described later.

(Conversion time difference)
Apart from the explanation of FIGS. 5 and 6, the difference between the virtual time and the real time will be described once. The difference obtained by subtracting the real time from the virtual time is called the "conversion time difference". In the above example, the abnormality sign diagnosis device 1 sets the conversion time difference of each step as "(PR1, PR2, PR3) = (0 seconds, 20 seconds, 40 seconds)". The conversion time difference of any one step is preferably "0 seconds". Further, it is preferable that the conversion time difference between the other two steps is not the same, and further, such as "(0 seconds, 10 seconds, 20 seconds)" and "(0 seconds, 20 seconds, 40 seconds)". , It is more desirable that the intervals are even.

However, if, for example, the interval between the real time t1 and the next real time t2 is 60 seconds, and the conversion time difference becomes "(0 seconds, 40 seconds, 80 seconds)", the virtual time of all processes becomes real. It cannot fit between the time t1 and the next real time t2. Therefore, in general, the abnormality sign diagnosis device 1 sets the conversion time difference to "(PR1, PR2, PR3, ...) = (0, p / q, 2p / q, 3p / q, ...)". Is most preferable. Here, p is a real-time interval and q is the number of parallel processes. When p = 60 seconds and q = 3 are satisfied, the most preferable conversion time difference is "(PR1, PR2, PR3) = (0 seconds, 20 seconds, 40 seconds)" as described above. This conversion time difference is called "standard time difference". The above-mentioned delay is a result of the abnormality sign diagnosis device 1 applying the standard time difference.

Returning to the description of FIGS. 5 and 6, for example, the following can be further understood.
The real time "t2" for the process PR1 is replaced with the virtual time "t21". “T2” is “October 10, 2017, 10:00:00”. “T21” is also “October 10, 2017, 10:00:00”.

The real time "t2" for the process PR2 is replaced with the virtual time "t22". "T2" is "October 10, 2017 10:00:00", while "t22" is "October 10, 2017 10:01:30". The virtual time is 20 seconds behind the real time. This delay is the result of the abnormality sign diagnostic device 1 applying the standard time difference.

The real time "t2" for the process PR3 is replaced with the virtual time "t23". "T2" is "October 10, 2017 10:00:00", while "t23" is "October 10, 2017 10:01:40". The virtual time is 40 seconds behind the real time. This delay is also a result of the abnormality sign diagnosis device 1 applying the standard time difference.

The real time "t3" for the process PR1 is replaced with the virtual time "t31". "T3" is "October 10, 2017 10:02:00". "T31" is also "October 10, 2017 10:02:00".

The real time "t3" for the process PR2 is replaced with the virtual time "t32". “T3” is “10:02:00 on October 10, 2017”, while “t32” is “10:02:20 on October 10, 2017”. The virtual time is 20 seconds behind the real time. This delay is the result of the abnormality sign diagnostic device 1 applying the standard time difference.

The real time "t3" for the process PR3 is replaced with the virtual time "t33". "T3" is "October 10, 2017 10:02:00", while "t33" is "October 10, 2017 10:02:40". The virtual time is 40 seconds behind the real time. This delay is also a result of the abnormality sign diagnosis device 1 applying the standard time difference.

(Effect of introducing virtual time and virtual process)
The effect of introducing the virtual time and the virtual process will be described with reference to FIGS. 7 and 8. Now, it is assumed that the abnormality sign diagnosis device 1 has read the measured value information 31 of FIG. 5 according to the existing technique. Then, the abnormality sign diagnosis device 1 recognizes that there are three steps and that the same four types of time-series measured values exist for each step. That is, the abnormality sign diagnosis device 1 recognizes that the information as shown in the upper left FIG. 71, the middle left FIG. 72, and the lower left FIG. 73 of FIG. 7 exists. The results of the diagnosis of the measured values (clustering using adaptive resonance theory) during the diagnosis period by the abnormality sign diagnosis device 1 on the premise of the existence of such information are shown in the upper left figure 81, the middle left figure 82, and the lower left figure of FIG. It is 83. It should be noted that these figures show the simplest two-dimensional plane in the multidimensional space, exclusively for explanatory purposes. The horizontal axis is the flow rate FA of the raw material A, and the vertical axis is the flow rate FB of the raw material B (the same applies to FIG. 84 on the right, which will be described later).

The position and radius of the sphere indicating category 1 in the upper left figure 81 is not always the same as the position and radius of the sphere indicating category 1 in the middle left figure 82, and the position and radius of the sphere indicating category 1 in the lower left figure 83. It is not always the same as the radius. The same is true for categories 2-4. That is, even if the measured value “◎” of the process PR1, the measured value “○” of the process PR2, and the measured value “●” of the process PR3 all appear to be included in the same category, they are the same abnormality (or). It is unknown whether it belongs to the same unknown state).

On the other hand, it is assumed that the abnormality sign diagnosis device 1 has read the measured value information 31b after sorting in FIG. Then, the abnormality sign diagnosis device 1 recognizes that there is only one process and that there are four types of time-series measured values for that one process. That is, the abnormality sign diagnosis device 1 recognizes that the information as shown in FIG. 74 on the right of FIG. 7 exists. The virtual process is a process recognized by the abnormality sign diagnosis device 1 in this way.

At first glance, FIG. 74 on the right of FIG. 7 is similar to FIG. 71 in the upper left, FIG. 72 in the middle left, and FIG. 73 in the lower left. However, while each of the upper left figure 71, the left middle figure 72, and the lower left figure 73 has only the measured value of one step, the right figure 74 has the measured value of three steps. That is, for example, the right figure 74 has all the measured values at the time indicated by the broken line in the upper left figure 71, the one-dot chain line in the middle left figure 72, and the two-dot chain line in the lower left figure 73 (the amount of data on the time axis is). It has tripled). In addition, "*" in FIG. 74 on the right collectively represents "1", "2" and "3".

FIG. 84 on the right side of FIG. 8 shows the result of the abnormality sign diagnosis device 1 diagnosing the measured value during the diagnosis period on the premise of the existence of such information. In FIG. 84 on the right, the measured value “⊚” of the process PR1 during the diagnosis period, the measured value “◯” of the process PR2, and the measured value “●” of the process PR3 are drawn in the same space (plane). Here, these three points belong to the sphere indicating category 4. The user knows that an unknown state has occurred in three parallel processes. The sphere showing category 1 in FIG. 84 on the right of FIG. 8 includes all of the spheres showing category 1 in FIG. 81 in the upper left, FIG. 82 in the middle left, and FIG. 83 in the lower left. The same is true for categories 2-4.

(Clustering using adaptive resonance theory)
Clustering using adaptive resonance theory will be described with reference to FIG. The abnormality sign diagnosis device 1 draws points indicating multidimensional measured values in a multidimensional space. A virtual time is associated with the measured value, and the number of drawn points corresponds to the number of virtual times. The two-dimensional plane of FIG. 9 is also an example of the simplest explanatory purpose in the multidimensional space, and has the flow rate FA of the raw material A on the horizontal axis and the flow rate FB of the raw material B on the vertical axis. The abnormality sign diagnosis device 1 classifies a plurality of points indicating the measured values acquired in the normal time into category 1, category 2, and category 3. These points are either the measured value “⊚” acquired in the process PR1, the measured value “◯” acquired in the process PR2, or the measured value “●” acquired in the process PR3.

Even during the diagnosis period, the abnormality sign diagnosis device 1 draws points (⊚, ◯, and ●) indicating measured values in a multidimensional space. These points may belong to any of Category 1, Category 2 or Category 3, and may not belong to any of them. In the example of FIG. 9, ⊚, ◯, and ● are repeatedly drawn in this order as the virtual time elapses. These points belong to category 4, which was unknown during the normal period. The user knows that in the three parallel steps, a state has occurred in which both the flow rate of the raw material A and the flow rate of the raw material B are at high levels.

(Abnormality and contribution)
The degree of abnormality and the degree of contribution will be described with reference to FIG. The vertical axis, the horizontal axis, and categories 1 to 4 in FIG. 10 are the same as those in FIG. When the measured value ● 85 in the diagnosis period belongs to category 4, which is unknown in the normal period, the user wants to know the following.
-How abnormal is the measured value from which known category?
-Which of the multiple types of measured values (flow rate of raw material A, flow rate of raw material B, etc.) contributes most to the abnormality?

Therefore, first, the abnormality sign diagnosis device 1 performs the following processing to calculate the degree of abnormality.
(1) The abnormality sign diagnosis device 1 specifies the representative point ◇ 86a of category 1, the representative point ◇ 86b of category 2, and the representative point ◇ 86c of category 3. Here, the representative point (representative value) is, for example, the median value (maximum value-minimum value), the average value (center of gravity), etc. of the measured values belonging to each category.
(2) The abnormality sign diagnosis device 1 calculates the distance between each specified representative point ◇ and the measured value ● 85, and sets these as candidates for the degree of abnormality. The number of candidate anomalies matches the number of known categories.
(3) The abnormality sign diagnosis device 1 sets the smallest candidate of the calculated abnormality degree as the “abnormality degree”. In the example of FIG. 10, the distance between the measured value ● 85 and the representative point ◇ 86c of category 3 is the degree of abnormality 87.

Subsequently, the abnormality sign diagnosis device 1 performs the following processing and calculates the degree of contribution.
(4) The abnormality sign diagnosis device 1 decomposes the abnormality degree 87 into a component in the horizontal axis direction and a component in the vertical axis direction. Here, "decomposing into components in the axial direction" means obtaining the length obtained by projecting the degree of abnormality 87 on each axis.
(5) In the abnormality sign diagnosis device 1, the length 88a is the contribution of the flow rate of the raw material A, and the length 88b is the contribution of the flow rate of the raw material B.

In general, the degree of anomaly in multidimensional space can be projected onto any axis in multidimensional space. That is, the number of contributions corresponds to the degree of the multidimensional space. As described above, the measured value is represented as a measured value vector having the value of each axis in the multidimensional space as a component. Therefore, it can be said that the degree of abnormality is the magnitude of the vector (subtraction result vector) obtained by subtracting the position vector of the representative point ◇ of the category from the measured value vector. Further, the contribution can be said to be an arbitrary component of the subtraction result vector itself.

(Processing procedure)
The processing procedure will be described with reference to FIG. In the middle of the description, reference to FIGS. 12 and 13 and the like as appropriate. As a premise for starting the processing procedure, it is assumed that the auxiliary storage device 15 stores the measured value information 31 (FIG. 5) for the normal period and the diagnosis period.
In step S201, the preprocessing unit 21 of the abnormality sign diagnosis device 1 learns the normal category. Specifically, first, the preprocessing unit 21 reads out the measured value information 31 from the auxiliary storage device 15.

Second, the preprocessing unit 21 accepts the user to input the normal period via the input device 12. The normal period is the period during which the plant is known to be normal. If the period desired to be a normal period partially includes a period in which a certain process is known to be abnormal, the user inputs a period excluding that period as a normal period. For the sake of brevity, it is assumed that no period was excluded here.
Thirdly, the preprocessing unit 21 extracts the record corresponding to the normal period received in the “second” of the step 201 from the measured value information 31 read in the “first” of the step S201.

Fourth, the preprocessing unit 21 sorts the records extracted in the "third" of step S201 by the method described above, and creates the sorted measurement value information 31b (FIG. 6).
Fifth, the preprocessing unit 21 draws points indicating the measured values in the multidimensional space based on the rearranged measured value information 31b created in the “fourth” of step S201, and a plurality of spheres (clusters). To create. At this stage, the preprocessing unit 21 is drawing category 1, category 2, and category 3 in FIG. 9 in the multidimensional space.

In step S202, the preprocessing unit 21 acquires the measured value of the diagnosis period. Specifically, first, the preprocessing unit 21 accepts the user to input the diagnosis period via the input device 12. The diagnosis period is a period in which it is not known whether or not each step is normal, and is usually a period following the normal period.
Secondly, the preprocessing unit 21 extracts the record corresponding to the diagnosis period received in the “first” of the step 202 from the measured value information 31 read in the “first” of the step S201.

In step S203, the preprocessing unit 21 rearranges the measured values during the diagnosis period. Specifically, the preprocessing unit 21 sorts the records extracted in the “second” of step S202 by the method described above, and creates the sorted measurement value information 31b (FIG. 6). At this time, the preprocessing unit 21 associates a plurality of different virtual times with respect to the measured values acquired from each of the parallel processes and having the same real time (see FIG. 4). .. In FIG. 4, for example, the virtual times t11, t12, and t13 are included in a predetermined time width (time width between real times t1 and t2).

In step S204, the classification unit 22 of the abnormality sign diagnosis device 1 classifies the measured values during the diagnosis period. Specifically, first, the classification unit 22 draws a point indicating the measured value in the multidimensional space based on the rearranged measured value information 31b created in step S203. Then, the classification unit 22 extracts the “post-processing target point” from the drawn points according to any of the following criteria.
<Criteria 1> Of the drawn points, points that do not belong to any of the known categories 1, category 2, and category 3 are set as post-processing target points.
<Criteria 2> All drawn points are set as post-processing target points.

Second, the classification unit 22 displays the multidimensional space on the output device 13. FIG. 9 is an example when the multidimensional space is a two-dimensional plane. At this time, it is desirable that the classification unit 22 also displays the process to which the post-processing target point belongs and the context of the virtual time. In the example of FIG. 9, the classification unit 22 changes the shape of the points for each process (⊚, ◯ and ●), and draws an arrow from the point where the virtual time is early to the point where the virtual time is late.

In step S205, the post-processing unit 23 of the abnormality sign diagnosis device 1 calculates the degree of abnormality and the degree of contribution. Specifically, the post-processing unit 23 calculates the degree of abnormality and the degree of contribution for each post-processing target point by the method described above in the explanation of FIG.

In step S206, the post-processing unit 23 displays the abnormality degree / contribution degree information 32 (FIG. 12). Specifically, the post-processing unit 23 displays the abnormality degree / contribution degree information 32 on the output device 13 for each post-processing target point.
The real time and / or the virtual time is displayed in the measurement time column 91a.
In the category column 91b, the number of the category to which the post-processing target point belongs is displayed. When following Criterion 1, the numbers here are numbers other than 1, 2 and 3.
In the normal / abnormal column 91c, "normal" or "abnormal" is displayed. If Criterion 1 is followed, "abnormal" is always displayed here.

In the abnormality cause column 91d, the type of the measured value having the largest contribution to the degree of abnormality (the largest component of the subtraction result vector) is displayed. When, for example, "F1A" is displayed in the column, it is understood that an abnormality has occurred in the process PR1 and that the change in the flow rate of the raw material A from the past example is likely to be the cause of the abnormality. ..
The degree of abnormality is displayed in the degree of abnormality column 91e. The degree of anomaly is the square root of the sum of squares of each component in a different unit of the subtraction result vector. Therefore, the degree of anomaly does not have its own unit. “[-]” Indicates this.

In the contribution ranking column 92a, the ranking counted from the one with the largest contribution is displayed. When the number of dimensions (the number of types of measured values) in the multidimensional space is n, the order is 1, 2, 3, ..., N.
In the measurement point column 92b, the type of the measurement value indicated by the axis on which the degree of abnormality is projected is displayed.
The contribution degree is displayed in the contribution degree column 92c.
The measured value itself is shown in the measured value column 92d.
When, for example, 10 post-processing target points exist, the post-processing unit 23 displays the abnormality degree / contribution degree information 32 of FIG. 12 10 times in the order of earliest virtual time.

In step S207, the post-processing unit 23 displays the diagnosis result information 33 (FIG. 13). Specifically, first, the post-processing unit 23 displays the diagnosis result information 33 on the output device 13. The diagnosis result information 33 is a collection of the abnormality degree / contribution degree information 32 displayed for each post-processing target point in step S206.

The category column 93a, the normal / abnormality column 93b, and the abnormality cause column 93d of the diagnosis result information 33 correspond to the columns having the same name in the abnormality degree / contribution degree information 32 of FIG.
“Learning” or “diagnosis” is displayed in the learning / diagnosis column 93c. “Learning” indicates that the category of the record was created for the normal (learning) period. "Diagnosis" indicates that the category of the record was created for the diagnostic period.
In the confirmed person column 93e, the user name (which may be an identifier) that confirmed the contents of the record is displayed.
In the fixed date column 93f, the date when the content of the record is confirmed is displayed.

The record of the diagnosis result information 33 is divided into an upper part and a lower part with the double line 95 as a boundary. The records 94a to 94c in the upper part are for the user to confirm the category created in the normal (learning) period. "Normal" is always displayed in the normal / abnormal column 93b. In the abnormality cause column 93d, "-" (no data) is always displayed. First, the post-processing unit 23 displays the confirmed person column 93e and the confirmed date column 93f of these records as blanks. After visually recognizing these and confirming that there are three normal categories, the user inputs his / her own identifier in the confirmed person field 93e of these records via the input device 12, and the confirmed date field 93f. Enter the confirmation date in.

Each of the lower records 96a to 96d corresponds to each of the post-processing target points in the diagnosis period. First, the post-processing unit 23 displays the confirmed person column 93e and the confirmed date column 93f of these records as blanks, and displays "diagnosis" in the learning / diagnosis column 93c. When the criterion 2 is followed, the post-processing unit 23 displays "normal" or "abnormal" in the normal / abnormal column 93b of these records.

When the post-processing unit 23 displays "abnormal" in the normal / abnormal column 93b, the post-processing unit 23 displays a number other than "1", "2" and "3" in the category column 93a of the record, and displays the abnormal cause column 93d. Displays the type of measurement value that contributes the most to. When the post-processing unit 23 displays "normal" in the normal / abnormal column 93b, "1", "2" or "3" is displayed in the category column 93a of the record, and "1", "2" or "3" is displayed in the abnormal cause column 93d. -”Is displayed.

After visually recognizing the records 96a to 96d, the user inputs his / her own identifier in the confirmed person field 93e of these records via the input device 12, and inputs the confirmed date in the confirmed date field 93f. The user can change the data displayed in the normal / abnormal column 93b and the abnormal cause column 93d as necessary, as shown in the following example.

<Example 1: Record 96b>
The post-processing unit 23 displayed "normal" in the normal / abnormal column 93b. However, the user wanted to put this diagnostic result on hold. At this time, the post-processing unit 23 accepts the user to change "normal" to "pending".

<Example 2: Record 96d>
The post-processing unit 23 displayed "abnormal" in the normal / abnormal column 93b. However, the user has determined that this diagnostic result is not due to a defect in the virtual process, but to a change in the environment in which the virtual process is located. At this time, the post-processing unit 23 accepts the user to change "normal" to "new" and delete the data displayed in the abnormality cause column 93d. Here, "new" means "normal" in a new category.

In addition to Examples 1 and 2, if the measured value is not included in any of the known categories, the post-processing unit 23 may first display "Caution required" in the normal / abnormal column 93b. Then, the user may change "attention required" to "new", "normal" or "abnormal".

The explanation of the screen display has become long, but I will return to the processing procedure. Second, the post-processing unit 23 accepts the user to perform an arbitrary operation (for example, pressing an "end" button (not shown)) via the input device 12.
Third, the post-processing unit 23 stores the diagnosis result information 33 after receiving the input of the identifier of the confirmed person and other changes in the auxiliary storage device 15, and ends the processing procedure.

(Effect of this embodiment)
The effects of the abnormality sign diagnosis device of this embodiment are as follows.
(1) The abnormality sign diagnosis device can classify abnormalities of the same cause into the same category for a plant composed of a plurality of parallel processes.
(2) The abnormality sign diagnosis device enables maintenance personnel to efficiently diagnose a plurality of parallel processes.
(3) The abnormality sign diagnosis device makes it easy for maintenance personnel to take measures to return the parallel process to the normal state.
(4) The abnormality sign diagnosis device can quantitatively show the basis of the diagnosis result to the maintenance staff.
(5) The abnormality sign diagnosis device can show the diagnosis result to the maintenance staff in a list format.

The present invention is not limited to the above-described embodiment, and includes various modifications. For example, the above-described embodiment has been described in detail in order to explain the present invention in an easy-to-understand manner, and is not necessarily limited to the one including all the configurations described. Further, it is possible to replace a part of the configuration of one embodiment with the configuration of another embodiment, and it is also possible to add the configuration of another embodiment to the configuration of one embodiment. Further, it is possible to add / delete / replace a part of the configuration of each embodiment with another configuration.

Further, each of the above-mentioned configurations, functions, processing units, processing means and the like may be realized by hardware by designing a part or all of them by, for example, an integrated circuit. Further, each of the above configurations, functions, and the like may be realized by software by the processor interpreting and executing a program that realizes each function. Information such as programs, tables, and files that realize each function can be placed in a memory, a recording device such as a hard disk or SSD (Solid State Drive), or a recording medium such as an IC card, SD card, or DVD.
In addition, the control lines and information lines indicate what is considered necessary for explanation, and do not necessarily indicate all the control lines and information lines in the product. In practice, it can be considered that almost all configurations are interconnected.

1 Abnormality sign diagnosis device 2 Plant 3 Measurement / control device 4 Network 11 Central control device 12 Input device 13 Output device 14 Main storage device 15 Auxiliary storage device 16 Communication device 21 Pre-processing unit 22 Classification unit 23 Post-processing unit 31 Measurement value information 32 Abnormality / contribution information 33 Diagnosis result information

Claims (6)

The measured values of a plurality of parallel processes in which the same type of device performs the same type of processing are acquired in chronological order for each of the parallel processes.
Among the measured values acquired from each of the parallel processes, a preprocessing unit that associates a plurality of different virtual times included in a predetermined time width with respect to those having the same actual acquired time, and a preprocessing unit.
A classification unit that considers that the measured value was acquired at the virtual time, and
An abnormality sign diagnostic device characterized by being equipped with. The pretreatment unit
The real time of a parallel process is used as the virtual time as it is.
To associate the virtual time of all parallel processes other than the parallel process with the time zone after the real time of the parallel process and before the next real time.
The abnormality sign diagnostic apparatus according to claim 1 . In a multidimensional space having various measured values as axes, it is provided with a post-processing unit that decomposes the distance between the representative value of the past measured value and the measured value at the time of diagnosis into components for any axis of the multidimensional space. matter,
The abnormality sign diagnostic apparatus according to claim 2 . The classification unit
Clustering using adaptive resonance theory
Displaying time-series measurements of multiple parallel processes in which the same type of equipment performs the same type of processing in the same space,
The abnormality sign diagnostic apparatus according to claim 3 . The pretreatment part of the abnormality sign diagnosis device is
The measured values of a plurality of parallel processes in which the same type of device performs the same type of processing are acquired in chronological order for each of the parallel processes.
Among the measured values acquired from each of the parallel processes, those having the same actual time actually acquired are associated with a plurality of different virtual times included in a predetermined time width.
The classification unit of the abnormality sign diagnosis device is
Assuming that the measured value was acquired at the virtual time ,
A method for diagnosing an abnormality sign of an abnormality sign diagnosis device, which comprises. For the pretreatment part of the abnormality sign diagnosis device
The measured values of a plurality of parallel processes in which the same type of device performs the same type of processing are acquired in chronological order for each of the parallel processes.
Among the measured values acquired from each of the parallel processes, those having the same actual time actually acquired are subjected to a process of associating a plurality of different virtual times included in a predetermined time width .
For the classification unit of the abnormality sign diagnosis device
To execute a process that considers the measured value to have been acquired at the virtual time .
An abnormality sign diagnosis program for operating the abnormality sign diagnosis device.

Images