WO2021220323A1

WO2021220323A1 - State determination device

Info

Publication number: WO2021220323A1
Application number: PCT/JP2020/017892
Authority: WO
Inventors: 健太霜田; 洋平上野; 健今井
Original assignee: 三菱電機株式会社
Priority date: 2020-04-27
Filing date: 2020-04-27
Publication date: 2021-11-04
Also published as: JP7118313B2; JPWO2021220323A1

Abstract

A state inference device (100, 300, 400, 500), in a processing network (60) including a plurality of processing systems (10, 21-24, 30, 41-44, 50) that carry out processing of an object to be processed, comprises: a sensor (S) that detects, as pieces of measured value information (J1), the respective states of the processing systems (10, 21-24, 30, 41-44, 50); and a determination unit (70) that carries out determination control for determining the respective states of the processing systems (10, 21-24, 30, 41-44, 50) on the basis of the detected pieces of measured value information (J1), wherein in the determination control, the determination unit (70) uses a reference determination formula f for determining the states of the processing systems (10, 21-24, 30, 41-44, 50) on the basis of the correlation between the respective pieces of measured value information J1 of the processing systems (10, 21-24, 30, 41-44, 50).

Description

State judgment device

The present application relates to a state determination device.

Treatment facilities for water and sewage, chemicals, steelmaking, etc. are composed of a series of multiple treatment systems that perform predetermined treatment. In general, the processing in each processing system affects the processing before and after each other, so a central monitoring control device is introduced to monitor and collectively control the processing status in the entire processing facility.

A state determination device exists to support the operation of such a central monitoring and control device.
In the state determination device, for example, data in which the state information of the processing system in which the equipment failure has occurred and the information indicating the behavior of the measured values of the processing system before and after the failure is registered is registered in advance. Then, when the measured value shows the behavior corresponding to the data, it is determined that an abnormality has occurred in the processing system, and the operator is notified.
However, when the information on the operation of the device is scarce, such as a newly introduced device, it is not possible to register in advance the data showing the behavior of the measured value when an abnormality occurs in the central monitoring control device, and it is related to sufficient operation. There is a problem that it is not possible to determine an abnormality in the processing system until the information is accumulated.

In response to these problems, it is possible to monitor the status of processing systems that have equipment that is similar to newly introduced equipment and has poor information on operation using data from equipment that has a track record of operation. A device abnormality monitoring system as such a state determination device is disclosed.
That is, the conventional device abnormality monitoring system creates an individual determination model for each of a plurality of existing similar devices. Next, a meta-prediction model is created in which the coefficients and intercepts of these individual prediction models are predicted from the feature item values of each device. Then, from this meta prediction model, a prediction model dedicated to the target device is generated. Using this determination model, the determination module monitors the state of the device and detects an abnormality (see, for example, Patent Document 1).

Japanese Unexamined Patent Publication No. 2010-287011 (paragraph [0038], FIG. 1)

The state determination device described in Patent Document 1 uses a prediction model dedicated to each processing system. Therefore, it is effective when only the judgment value of a single treatment system is judged, while in a treatment facility composed of a plurality of treatment systems such as a water and sewage treatment facility, the measured values of each of these treatment systems are combined. When it is necessary to determine the state, the following difficulties arise.
That is, in Patent Document 1, it is determined whether or not an abnormality has occurred in a specific processing system among a plurality of processing systems based only on the measured values of the specific processing system regardless of the behavior of the measured values of the other processing systems. I do. Therefore, when the measured value of a specific processing system shows an unusual behavior, for example, if this behavior is a fluctuation within the normal range, it may be that this different behavior is caused by the normal processing, or the sensor fails. There was a problem that it was not possible to determine whether the cause was due to or another cause.

In addition, when determining the entire multiple processing systems, a prediction model showing the behavior of all measured values, including sensor failures in each processing system, must be registered in advance for each processing system, which requires design. The problem of increased load remains.

The present application discloses a technique for solving the above-mentioned problems, and in a processing network composed of a plurality of processing systems, it is possible to determine the state of the processing system with high accuracy and design it. It is an object of the present invention to provide a state determination device capable of reducing such a load.

The state determination device disclosed in the present application is
In a processing network composed of a plurality of processing systems for processing an object to be processed
A detection unit that detects the state of each of the processing systems as measured value information,
A determination unit that performs determination control for determining the state of the processing system based on the detected measured value information is provided.
In the determination control, the determination unit
A reference determination formula for determining the state of the processing system based on the correlation between the measured value information of each processing system is used.
It is a thing.

According to the state determination device disclosed in the present application, a state determination device capable of performing the state determination of the processing system with high accuracy and reducing the load on the design can be obtained.

It is a block diagram which shows the schematic structure of the state determination apparatus according to Embodiment 1. FIG. FIG. 5 is a schematic diagram of a processing network in which a state determination device according to the first embodiment determines a state. It is a figure which schematically represented the recording method at the time of recording the configuration information of the processing network by Embodiment 1 in a network model information database. It is a block diagram which shows the schematic structure of the state determination apparatus according to Embodiment 3. It is a block diagram which shows the schematic structure of the state determination apparatus according to Embodiment 4. It is a block diagram which shows the schematic structure of the state determination apparatus according to Embodiment 5.

Embodiment 1.
FIG. 1 is a block diagram showing a schematic configuration of the state determination device 100 according to the first embodiment.
FIG. 2 is a schematic view of a processing network 60 in which the state determination device 100 according to the first embodiment determines a state.
First, the processing network 60 in which the state determination device 100 of the present embodiment determines the state will be described with reference to FIG.
As shown in FIG. 2, for example, in a water and sewage treatment facility, a treatment system such as a water tank (first treatment system 10, second treatment system) that performs a set treatment such as purification treatment on an object to be treated such as raw water. 21 to 24, a third processing system 30, and a plurality of fourth processing systems 41 to 44) are provided to form a processing network 60.

First, the object to be processed sent from the upstream processing system (not shown) is sent to the parallel processing system 20 in the subsequent stage after performing the first processing set in the first processing system 10.

The parallel processing system 20 is configured by connecting four

second processing systems

21, 22, 23, and 24 in parallel by a water channel or the like. Then, these

second processing systems

21, 22, 23, 24 perform the set second processing in parallel with respect to the object to be processed that has been sent from the first processing system 10 in the previous stage and has been diverted. .. The

second processing systems

21, 22, 23, and 24 are used or not used properly depending on the operating conditions.
The object to be processed is sent to the third processing system 30 in the subsequent stage after the second processing set in the parallel processing system 20 is performed.

In the third processing system 30, the objects to be processed that have been processed in parallel in the

second processing systems

21, 22, 23, and 24 are collected.
The object to be processed is sent to the parallel processing system 40 in the subsequent stage after the third processing set in the third processing system 30 is performed.

Similar to the parallel processing system 20, the parallel processing system 40 is configured by connecting four

fourth processing systems

41, 42, 43, 44 in parallel by a water channel or the like. Then, these

fourth processing systems

41, 42, 43, 44 carry out the set fourth processing in parallel with respect to the object to be processed that has been sent from the third processing system 30 in the previous stage and has been diverted. .. As with the parallel processing system 20, the

fourth processing systems

41, 42, 43, and 44 are used or not used properly depending on the operating conditions.
The object to be processed is sent to the fifth processing system 50 in the subsequent stage after the fourth processing set in the parallel processing system 40 is performed.
In this way, the treatment of the object to be treated is performed stepwise by the first, second, third, fourth, and

fifth treatment systems

10, 21 to 24, 30, 41 to 44, and 50.

In addition, each processing system is equipped with a sensor as a detection unit for confirming the state of the processing system.
In the present embodiment, the second processing system 21 of the parallel processing system 20 has a sensor 21S, the second processing system 22 has a sensor 22S, the second processing system 23 has a sensor 23S, and the second processing system 24 has a sensor 24S. Is installed.
Further, a sensor 41S is installed in the fourth processing system 41 of the parallel processing system 40, a sensor 42S is installed in the fourth processing system 42, a sensor 43S is installed in the fourth processing system 43, and a sensor 44S is installed in the fourth processing system 44. There is.
Hereinafter, when the

sensors

21S, 22S, 23S, 24S, 41S, 42S, 43S, and 44S are collectively referred to, they are referred to as the sensor S.

The installation position where the sensor S is installed in the processing process may be determined according to the information that the sensor S wants to measure. In the parallel processing systems 20 and 40 shown in FIG. 2, sensors 21S to 24S and 41S to 44S are installed on the input side of the object to be processed in the parallel processing systems 20 and 40. Further, in the fifth processing system 50, the sensor S is installed on the output side of the object to be processed.
Further, the type of the sensor S is not one type, and a plurality of types may be installed.

Hereinafter, the configuration of the state determination device 100 of the present embodiment will be described with reference to FIG.
The state determination device 100 is a

treatment system

10, 21 to 24, 30, 41 to 44, 50 based on the measured values of the sensors S attached to each device and equipment of each treatment system constituting the water and sewage treatment facility and the like. It determines the state of.

As shown in FIG. 1, the state determination device 100 includes a determination unit 70, an input / output device 1, a sensor S, and a collection device 2.
The sensor S detects the states of the

processing systems

10, 21 to 24, 30, 41 to 44, and 50 as the measured value information J1. The measured value information J1 may be, for example, the water quality of the raw water as the object to be measured, or the voltage in the equipment and facilities of the treatment system.

The detected measured value information J1 is collected by the collecting device 2 and transmitted to the determination unit 70.
The determination unit 70 performs determination control for determining the state of each of the

processing systems

10, 21 to 24, 30, 41 to 44, and 50 described below based on the measured value information J1, and determines the determination result in the state information J2. Output as. The state information J2 is transmitted to the operator through the input / output device 1.

Hereinafter, the configuration of the determination unit 70 and the determination control performed by the determination unit 70 will be described.
As shown in FIG. 1, the determination unit 70 includes a measurement information database 71 (hereinafter referred to as a measurement information DB), a reference determination type information database 72 (hereinafter referred to as a judgment type information DB), and processing process information. It includes a network model information database 73 (hereinafter referred to as N model information DB), a determination formula creation unit 74, and a state determination unit 75.

In the measurement information DB 71, the type of the sensor S and the measurement value information J1 detected by the sensor S are recorded for each measurement time.

The N model information DB 73 electronically records configuration information such as connection relationships of the

processing systems

10, 21 to 24, 30, 41 to 44, and 50 of the processing network 60.
FIG. 3 is a diagram schematically showing a recording method when recording the configuration information of the processing network 60 shown in FIG. 2 in the N model information DB 73.

As shown in FIG. 3, the processing process name 73A, the parallel processing process name 73B, the pre-stage processing process name 73C, the post-stage processing process name 73D, and the incidental measurement value name 73E are registered in the N model information DB 73.
The processing process name 73A indicates a processing system of each stage in the processing network 60, the first processing system 10 is processing 1, the second processing systems 21 to 24 in the subsequent stage are processing 2, and the third processing system in the subsequent stage. The processing system 30 is shown as processing 3, ....
When a plurality of second processing systems 21 to 24 exist in one stage as in the parallel processing system 20, they are collectively shown as processing 2.

The parallel processing process name 73B is used in the

respective processing systems

21, 22, 34, and 24 when a plurality of

processing systems

21, 22, 23, and 24 exist in parallel as in the parallel processing system 20 of FIG. A name unique to the second process to be performed (processes 2-1, 2-2, 2-3, 2-4 in FIG. 3) is recorded.

The pre-stage processing process name 73C indicates the name of the processing system on the front stage side which is the transmission source to the processing system recorded in the parallel processing process name 73B, and the post-stage processing process name 73D is the name of the processing system on the rear stage side which is the destination. It indicates the processing process name.
For example, since the pre-stage processing system that is the source of the second processing (processing 2-1, 2-2, 2-3, 2-4) of the parallel processing system 20 is the first processing system 10, the pre-stage processing step Process 1 indicating the first process is recorded under the name 73C. Further, since the subsequent processing system that is the destination of the second processing (processing 2-1, 2-2, 2-3, 2-4) of the parallel processing system 20 is the third processing system 30, the name of the subsequent processing process. In 73D, the process 3 indicating the third process is recorded. In this way, by combining the processing process name 73A with the pre-stage processing process name 73C and the post-stage processing process name 73D, the processing network 60 as shown in FIG. 2 can be described in the form of a network graph.
In this way, when the processing for the object to be processed is performed stepwise by a plurality of processing systems, the N model information DB 73 indicates the processing order of each processing system that is performed stepwise.

Further, the incidental measurement value name 73E records the name of the sensor S attached to the processing systems 21 to 24, 41 to 44, and 50 recorded in the parallel processing process name 73B, and the second processing system 21,

Sensors

21S, 22S, 23S, and 24S correspond to 22, 23, and 24, respectively.

The reference determination formula f used in the determination control is recorded in the determination expression information DB 72.
This reference determination formula f is based on the correlation between the measured value information J1 detected in each of the processing systems 21 to 24, 41 to 44, and 50, respectively, in each of the

processing systems

10, 21 to 24, 30, 41 to 44, It determines the state of 50.
For example, the measured value information J1 detected in the second processing system 21 and the measured value information J1 detected in the second processing system 22 are used as input values of the reference determination formula f. Then, the reference determination formula f determines the state of the second processing system 21 or the second processing system 22 based on the correlation between these two measured value information J1, and outputs the determination result as a calculated value.

Further, the reference determination formula f is configured to be common to each combination of parallel processing systems in the parallel processing systems 20 and 40.
For example, in the case where four of the

second processing systems

21, 22, 23, and 24 exist in parallel as in the parallel processing system 20, for example, the combination of the processing systems based on the second processing system 21 is "the first. 2 processing system 21-2nd processing system 22 "," 2nd processing system 21-2nd processing system 23 "," 2nd processing system 21-2nd processing system 23 "," 2nd processing system 21-2nd processing System 24 ". The reference determination formula f is configured to be one determination formula commonly used for these four combinations.
Further, the reference determination formula f is created by the determination formula creation unit 74 as described below.

The determination formula creation unit 74 of each processing system 21 to 24, 41 to 44, 50 based on the connection relationship of each

processing system

10, 21 to 24, 30, 41 to 44, 50 shown in the N model information DB 73. Some combinations or all combinations are arbitrarily selected and set. Then, the determination expression creating unit 74 creates one reference determination expression f commonly used for each set combination and records it in the determination expression information DB 72.

The state determination unit 75 calls the corresponding reference determination expression f from the determination expression information DB 72 through the determination expression creation unit 74. For example, when the state determination is performed on the second processing system 21 of the parallel processing system 20, the reference determination formula f common to each combination with the second processing system 21 as a reference is called.
With respect to the called reference determination expression f, the state determination unit 75 determines the state of the processing system by using the measurement value information J1 corresponding to each combination as the input value of the reference determination expression f. For example, the reference determination formula f is based on the second processing system 21, "second processing system 21-2nd processing system 22", "second processing system 21-2nd processing system 23", and "second processing". When configured in common for all combinations of "system 21-2nd processing system 23" and "second processing system 21-2nd processing system 24", all detected in the

sensors

21S, 22S, 23S, 24S. The measured value information J1 is used as an input value. Further, for example, when the reference determination formula f is configured corresponding to the combination of "second processing system 21-2nd processing system 22", the measured value information J1 of the

sensors

21S and 22S is used as an input value.

The state determination unit 75 uses the reference determination formula f in this way to perform determination control for determining the state of the processing system by comparing the correlation between the measurement value information J1 of each sensor S.
For example, it is assumed that the measured value information J1 of one sensor S fluctuates within the fluctuation range of the steady state of the processing system.
In this case, when the comparison value indicating the correlation between the measured value information J1 of one sensor S and the measured value information J1 of the other sensor S is remarkably large, the state determination unit 75 detects the sensor S in the sensor S in which the fluctuation occurs. Determine that a failure has occurred.

According to this method, by comparing the correlation of the behavior between the measured value information J1 of the sensor S, the measured value within the steady processing range by the behavior pattern of the unknown measured value information J1 or the operation by the operator or the like. Even when the information J1 changes, it is possible to accurately determine the state of the processing system such as steady state or abnormality with high accuracy. In this way, the sensor S that behaves differently from the other sensors S can be identified, and the operator can be accurately notified, for example, that a failure has occurred.

Further, the determination result by the state determination unit 75 is registered in the determination expression information DB 72 through the determination expression creation unit 74. In this way, the determination formula information DB 72 records the determination result, which is information on whether the state of the processing system is steady or abnormal.
Further, the determination formula information DB 72 may record not only the determination result by the determination formula creation unit 74 but also the state of the processing system that can be grasped in advance. For example, information such as classification of the operation mode of the processing system and whether the processing system is under inspection is registered.
The registration method includes a success / failure judgment based on a threshold value, a judgment based on a difference evaluation from a definition formula, and the like, and is recorded in the judgment formula information DB 72 in the form of a threshold value or a definition formula.

As described above, the determination formula creation unit 74 creates the reference determination formula f common to each combination of the processing systems based on the connection relationship of the processing systems shown in the N model information DB 73. Here, the determination formula creation unit 74 may create the reference determination formula f based on the determination result of the processing system registered in the determination formula information DB 72 as described above. For example, when it is determined that the sensor S of the second processing system 22 is out of order, a reference determination formula f of a combination that does not use the measurement value information J1 of the sensor 22S is created. In this way, the determination formula creation unit 74 can dynamically create an appropriate reference determination formula f according to the actual operating state of each processing system.

Further, in the above, the reference determination formula f is configured to be common to each combination of the parallel processing systems in the parallel processing systems 20 and 40, but the present invention is not limited to this. For example, when performing determination control across the parallel processing system 40 and the fifth processing system 50, the reference determination formula f is common to the fourth processing system and the fifth processing system 50 in the parallel processing system 40. Should be configured. Here, the parallel processing system 40 has four measured value information J1 in the four fourth processing systems 41 to 44, and the fifth processing system 50 has one measured value information J1. In this case, as the measured value information J1 in the parallel processing system 40, the maximum value, the minimum value, the average value, and the like of the four measured value information J1 in the parallel processing system 40 are used. In this case, the type (for example, maximum value) of the measured value information J1 used as the measured value information J1 in the parallel processing system 40 and the selection method thereof are recorded in the determination formula information DB 72.

The state determination device of the present embodiment configured as described above is
In a processing network having a plurality of processing systems for processing an object to be processed
A detection unit that detects the state of each of the processing systems as measured value information,
A determination unit that performs determination control for determining the state of the processing system based on the detected measured value information is provided.
In the determination control, the determination unit
A reference determination formula for determining the state of the processing system based on the correlation between the measured value information of each processing system is used.
It is a thing.

Generally, when detecting a sensor failure as an abnormality of a processing system, for example, a method is used in which a range to be determined as "normal" is determined in advance for each sensor, and when the range is deviated, a failure is determined. .. However, it is difficult to detect a new abnormal state that has not been registered in advance in such a method of individually determining each processing system, and the range of "normal" changes depending on the operation method of the operator. In some cases, it becomes difficult to define the state of the processing system. Further, in a processing facility composed of a plurality of processing systems, the processing of each processing system influences each other. Therefore, in this case as well, it is difficult to accurately determine the state by the method of determining each processing system individually. ..

On the other hand, the state determination device of the present embodiment includes a detection unit that detects the state of each processing system as measured value information. Then, the determination unit uses a reference determination formula for determining the state of the processing system based on the correlation between the measurement value information of each processing system in the determination control.
In this way, the detection unit detects the state of the processing system, which changes according to the operating state of the processing system, as detection value information at any time. Then, by determining the state of the processing system based on the correlation between the detected value information that changes according to the operating state, it is based on the behavior of each measured value information regardless of the behavior of each measured value information. The state can be determined. In this way, the state determination can be performed with high accuracy.

For example, in water and sewage treatment facilities, a method of preparing multiple same treatment systems and using them properly depending on the situation is adopted from the viewpoint of increase / decrease in the amount of purified water, equipment maintenance, and the like. By applying the state determination device of the present embodiment to such a processing facility, when the sensor outputs a value different from the normal value due to a sensor failure in one processing system, the same type of sensor measurement value in each processing system By comparing these with each other, it is possible to determine whether a sensor failure has occurred or an abnormality related to the entire processing facility has occurred.

Further, the state determination device of the present embodiment configured as described above is
The determination unit
It has processing process information indicating the connection relationship of a plurality of the processing systems in the processing network, and has processing process information.
In the determination control, a combination of a plurality of the processing systems is set based on the connection relationship of the processing systems shown in the processing process information, and the reference commonly used for each of the set combinations. Create a judgment formula,
It is a thing.

In this way, the reference judgment formula used in the judgment control is configured to be commonly used for the mutual combination of the set processing systems. Therefore, it is not necessary to create and record a determination formula for each processing system or for all combinations, and it is only necessary to record one reference determination formula common to each combination. As a result, the load on the design can be reduced, the cost can be reduced, and the processing load on the determination unit can be reduced.

Further, the determination unit has processing process information indicating the connection relationship of a plurality of processing systems in the processing network. Then, the mutual combination of the processing systems is set based on this connection relationship, and the reference determination formula is created. In this way, the determination unit sets the combination in the determination control based on the connection relationship shown in the processing process information, thereby determining the combination according to the actual system network state and the reference based on this combination. Judgment formula can be created. Therefore, for example, when there is a processing system that is in the process of being inspected and is not in operation, a reference judgment formula excluding this processing system can be dynamically created and judgment control can be performed, so that the processing load in the judgment unit can be performed. Can be reduced and the state can be determined more accurately.

Further, the state determination device of the present embodiment configured as described above is
The processing network is configured to have a parallel processing system that processes the object to be processed in parallel by the plurality of processing systems.
In the determination control, the determination unit
Based on the connection relationship shown in the processing process information, the reference determination formula commonly used for each combination of the processing systems having a parallel relationship in the parallel processing system is created.
It is a thing.

In this way, the judgment unit creates a reference judgment formula that is commonly used for each combination of each processing system in the parallel processing system. In particular, in a parallel processing system that processes an object to be processed in parallel, the state of the processing system is likely to be reflected in the correlation between the measured value information. Therefore, by performing the determination control using the above-mentioned reference determination formula in the parallel processing system, it is possible to accurately determine the state of the processing system.

Embodiment 2.
Hereinafter, the second embodiment of the present application will be described with reference to the parts different from the first embodiment. The same parts as those in the first embodiment are designated by the same reference numerals, and the description thereof will be omitted.
In the present embodiment, an abnormality of a complicated factor such as an abnormality that occurs in the processing system infrequently and it is difficult to grasp the cause of the abnormality occurs in the processing system, or an abnormality finally occurs due to a combination of a plurality of events. Will be described when the above occurs in the processing system.

When an abnormality occurs in the processing system as described above, the operator needs to identify the location causing the abnormality in the processing network 60 in order to accurately grasp the state. In identifying the cause location, the change in the behavior of the measured value information J1 in the processing system in the first stage or the second stage of the abnormality processing system is investigated.
Then, the processing system that is the cause of the first stage and the second stage of the abnormality processing system is specified, and the behavior of the measured value information J1 in the processing system is extracted and the analysis is advanced. The details will be described below.

It is assumed that the measured value information J1 value of the sensor 50S of the fifth processing system 50 shown in FIG. 2 is in an abnormal state, and the fifth processing system 50 is an abnormal processing system.
The state determination unit 75 obtains the measured value information J1 of the processing system in the first stage or the second stage of the fifth processing system 50 based on the processing order information of each processing system that is performed stepwise, which is registered in the N model information DB 73. Extract. The fifth processing system 50 does not have a processing system on the rear stage side, and the fourth processing systems 41 to 44 in the parallel processing system 40 correspond to the processing system in the front stage.

Therefore, the state determination unit 75 extracts the measurement value information J1 of the sensors 41S to 44S of the fourth processing systems 41 to 44. Further, the state determination unit 75 extracts the measurement value information J1 of the processing system in the previous stage of the parallel processing system 40. Since the first and

third processing systems

10 and 30 do not have the measured value information J1, the state determination unit 75 extracts the measured value information J1 of the sensors 21S to 24S of the parallel processing system 20.

The state determination unit 75 is registered in the reference determination formula information database 72 for the measured value information J1 of the parallel processing systems 20 and 40 on the front stage side of the fifth processing system, which is the abnormality processing system, extracted in this way. Using the reference determination formula f, determination control is performed from the parallel processing system 40 to the parallel processing system 20.
By this determination control, for example, the fluctuation of the value of the measured value information J1 of the sensors 41S to 44S in the parallel processing system 40 is different from the normal value, and the value of the measured value information J1 of the sensors 21S to 24S in the parallel processing system 20 is normal. It is determined that it is within the range. In this case, the state determination unit 75 determines that the location (processing system) that caused the abnormality in the fifth processing system 50 is the parallel processing system 20 or the third processing system 30.

The state determination device of the present embodiment configured as described above is
The processing on the object to be processed is carried out stepwise by the plurality of the processing systems.
The processing process information further indicates the processing order of the processing system to be performed stepwise.
In the determination control, the determination unit
When an abnormal processing system, which is the processing system determined to be in an abnormal state, is detected, the processing belonging to the front stage side or the rear stage side of the abnormal processing system is based on the processing order shown in the processing process information. Judging the system status,
It is a thing.

In this way, the processing process information indicates the processing order of the processing system that is performed step by step. Then, the determination unit determines the state of the processing system belonging to the stage on the front stage side or the rear stage side of the abnormality processing system based on the processing order shown in the processing process information. As a result, in a complicated processing network that processes the object to be processed step by step, even if the occurrence frequency is low and an abnormality of a complicated factor occurs, the steps are performed according to the processing order shown in the processing process information. By investigating the presence or absence of anomalies, it is possible to identify the cause of the anomaly in the processing network. As a result, stable operation of the processing facility can be ensured.

Embodiment 3.
Hereinafter, the third embodiment of the present application will be described with reference to the parts different from those of the first and second embodiments.
In the second embodiment, it has been described that a portion causing an abnormality is specified in the processing network 60. In the present embodiment, when the process of occurrence of an abnormality and the behavior pattern of the measured value information J1 in the processing system of the specified location can be defined as a result of identifying the cause location, the determination formula information DB 72 is newly added. Add a statistical processing formula. As a result, even when the same abnormality occurs next time, it is possible to easily identify the cause. The details will be described below.

FIG. 4 is a block diagram showing a schematic configuration of the state determination device 300 according to the third embodiment.
The state determination device 300 in the present embodiment is obtained by adding the determination formula creation unit 374 to the determination unit 70 of the state determination device 100 shown in the first embodiment.
Further, in the present embodiment, a plurality of statistical processing formulas ft for deriving the evaluation value of the measured value information J1 are recorded in the determination formula information DB 72. The evaluation value of the measured value information J1 is the maximum value, the minimum value, the average value, the gain range, the regression equation, and the maximum value, the minimum value, the average value, the gain range, and the regression equation of the measured value information J1 when a plurality of the measured value information J1 is detected within the set period. The correlation between the measured value information J1 and the like are the evaluation values. The formula for deriving each of these evaluation values is the statistical processing formula ft.

First, it is assumed that the state determination unit 75 detects an abnormality in a certain processing system in the determination control as shown in the first embodiment. Here, the period including the period in which this processing system is in an abnormal state is defined as the first period. Then, the period in which the processing system before the first period is in the normal state will be described as the second period.

In this determination control, the state determination unit 75 calls a desired statistical processing expression ft from among a plurality of statistical processing expression ft recorded in the determination expression information DB 72.
In the present embodiment, the state determination unit 75 uses a statistical processing formula ft for deriving the minimum value of the measured value information J1, a statistical processing formula ft for deriving the maximum value, and a statistical processing formula ft for deriving the average value. call.
Then, the state determination unit 75 sets the minimum value, maximum value, and average value (hereinafter, referred to as minimum value Min1, maximum value Max1, and average value Ave1) of the measured value information J1 within the first period into each statistical processing formula ft. Derived from.
Further, the state determination unit 75 sets the minimum value, maximum value, and average value (hereinafter, referred to as minimum value Min2, maximum value Max2, and average value Ave2) of the measured value information J1 within the second period into each statistical processing formula ft. Derived from.

Then, the state determination unit 75 compares these evaluation values between the first period and the second period.
That is, the state determination unit 75 compares the minimum value Min1 in the first period with the minimum value Min2 in the second period, compares the maximum value Max1 with the maximum value Max2, and compares the average value Ave1 with the average value Ave2. conduct.
Then, the state determination unit 75 performs determination control for determining the state of the processing system based on the evaluation value having the largest difference among the evaluation values of the evaluation values of the first period and the second period. For example, when the difference between the maximum value Max1 and the maximum value Max2 is the largest among the differences between the maximum value, the minimum value, and the average value, the maximum value Max1 and the maximum value Max2 in which the difference appears most are used. Judgment control is performed.

The determination formula creation unit 374 uses the statistical processing formula ft for deriving the maximum value with respect to the evaluation value (maximum value) of the measurement value information J1 that the state determination unit 75 has determined to be related to the occurrence of an abnormality. It is automatically recorded as a new reference judgment formula f in the judgment formula information DB 72.
Then, the state determination unit 75 uses the automatically updated new reference determination formula f as the new reference determination formula f used for the state determination of the processing system. As a result, the evaluation value (maximum value) at which the most abnormal sign appears in the processing system to be monitored can be used for the determination control, and the state determination of the processing system can be performed with higher accuracy.

In addition, the operator can grasp the signs of abnormality in the processing system to be monitored and perform detailed analysis by checking and correcting the contents of the newly updated standard judgment formula f that has been automatically updated. Further, even if the operator does not have specialized knowledge about data analysis work and statistical processing, the appropriate standard judgment formula f is automatically created and updated in this way easily, so that the processing facility can be used. Stable operation becomes possible.

Further, the state determination unit 75 may perform a state determination as described below.
The state determination unit 75 sets the first distribution information D1 of the measured value information J1 as the evaluation value within the first period, which is the abnormal period, and the evaluation value within the second period, which is the normal period, in the abnormality handling system in which the abnormality has occurred. The second distribution information D2 of the measured value information J1 is derived.
Further, the state determination unit 75 derives the third distribution information D3 of the measured value information J1 as an evaluation value within a set period in another processing system in which no abnormality has occurred in the processing network 60.

Then, the state determination unit 75 compares the third distribution information D3 in the processing system in which the abnormality has not occurred with at least one of the first distribution information D1 in the abnormal period or the second distribution information D2 in the normal period in the abnormal processing system.
As a result, based on at least one of the behavior of the measured value information J1 in the abnormality processing system during the abnormality period or the behavior immediately before the abnormality, the processing system in a state similar to these behaviors is in a state where no abnormality has occurred. Can be identified. By identifying and understanding the treatment systems that can cause abnormalities even in the normal state in this way, it is possible to take measures before abnormalities occur in each treatment system of the water and sewage treatment facility and perform appropriate maintenance treatment. It becomes.

The state determination device of the present embodiment configured as described above is
In the determination control, the notation determination unit
The first distribution information of a plurality of the measured value information detected in the first period including the period of the abnormal state in the abnormality handling system, and the detection in the second period prior to the first period. At least one of the second distribution information of the measured value information of the plurality of the abnormal processing systems, and
In each of the processing systems other than the abnormality processing system, the third distribution information of the plurality of the measured value information detected within the set period is derived.
Using at least one of the first distribution information and the second distribution information and the third distribution information, the state of each of the processing systems other than the abnormal processing system is determined.
It is a thing.

In this way, other processing systems that resemble the behavior of the measured value information before and after the occurrence of an abnormality in the abnormality processing system can be detected, so that, for example, appropriate maintenance processing of the water and sewage treatment facility becomes possible and processing. Stable operation of the facility can be ensured.

Further, the state determination device of the present embodiment configured as described above is
The determination unit
It has a plurality of statistical processing formulas for deriving an evaluation value based on a plurality of the measured value information detected within a set period.
In the determination control, the evaluation values corresponding to the statistical processing formulas are derived for each of the first period and the second period.
The statistical processing formula having the largest difference in the evaluation value between the first period and the second period is selected from a plurality of the statistical processing formulas, and the selected statistical processing formula is used as a new reference determination formula. Used as
It is a thing.

In this way, the judgment unit uses the statistical processing formula with the largest difference in evaluation values as a new standard judgment formula in judgment control. As a result, the evaluation value at which the most abnormal sign appears in the processing system to be monitored can be used for the determination control, and the state determination of the processing system can be performed with higher accuracy.

Embodiment 4.
Hereinafter, the fourth embodiment of the present application will be described with reference to the parts different from the first embodiment.
When a plurality of second processing systems 21 to 22 exist in parallel as in the parallel processing system 20, measurement between the second processing systems 21 to 22 in parallel due to the adjustment of the sensor S and the difference in civil engineering structure. There are cases where the behavior of the value information J1 is different. For example, when the water channel from the first treatment system 10 to the second treatment system 21 to 24 is branched due to physical characteristics such as an open channel, the flow rate flowing into the second treatment system 21 to 24 is completely divided into four equal parts. Instead, there are variations in each channel.

In these cases, the distribution and tendency of the measured value information J1 of each of the second processing systems 21 to 24 varies with respect to the common reference judgment formula f registered in the judgment formula information DB 72, so that the determination of the occurrence of an abnormality is the first. 2 It may be different for each processing system 21 to 24.
The state determination device 400 of the fourth embodiment is proposed as a solution to such a problem, and will be described below with reference to FIG.

FIG. 5 is a block diagram showing a schematic configuration of the state determination device 400 according to the fourth embodiment.
The state determination device 400 of the present embodiment adds a feature amount extraction unit 476 and a correction coefficient information database 477 (hereinafter referred to as a correction coefficient information DB) to the state determination device 100 shown in the first embodiment. It was done.
Here, the feature amount extraction unit 476 is the processing system in the parallel processing systems 20 and 40 based on the connection relationship of the

processing systems

10, 21 to 24, 30, 41 to 44, and 50 shown in the N model information DB 73. 21 to 24 and 41 to 44 are extracted. Then, the distribution and correlation between the measured value information J1 in parallel are compared, and the difference thereof is registered in the correction coefficient information DB 73 as the correction coefficient information. The details will be described below.

The feature amount extraction unit 476 uses the third of the input values of the reference determination formula f for the reference determination formula f registered in the judgment formula information DB 72, which uses the measurement value information J1 of the parallel processing systems 20 and 40, respectively. The distribution information D3 and the fourth distribution information D4 of the output value, which is the calculation result of the reference determination formula f, are calculated respectively.
Further, when the distribution tendency of the third distribution information D3 and the distribution tendency of the fourth distribution information D4 are different from each other, it is a judgment result of the judgment control between the parallel processing systems 21 to 24 and 41 to 44. A correction coefficient is created so that the detection frequency of abnormality occurrence is the same, and is recorded in the correction coefficient information DB 73.

When the measurement value information J1 to be judged belongs to the parallel processing systems 20 and 40, the judgment formula creation unit 74 corrects the reference judgment formula f from the correction coefficient recorded in the correction coefficient information DB477.

The state determination device of the present embodiment configured as described above is
The processing network is configured to have a parallel processing system that processes the object to be processed in parallel by the plurality of processing systems.
The determination unit
The processing of the third distribution information of the measured value information as the input value of the reference determination formula in the determination control and the fourth distribution information of the calculation result as the calculation value of the reference determination formula in the parallel processing system. Derived for each combination of systems,
Correction to correct the reference discriminant so that the determination result in the determination control is the same in each processing system in the parallel processing system based on the derived third distribution information and the fourth distribution information. Create a coefficient,
It is a thing.

In this way, the determination unit derives the distribution information of the input value of the reference determination expression and the distribution information of the calculation result for each combination of the parallel processing systems. Then, based on each of the derived input values and the distribution information of the calculation result, the reference determination formula is corrected so that the determination result in the determination control is the same in each of the processing systems in the parallel processing system.
As a result, if there is a variation in distribution or tendency between parallel systems, by automatically correcting this, the frequency of abnormal occurrences can be leveled, and over-detection and oversight of abnormal conditions can be suppressed, which is appropriate. It is possible to judge the state.

Embodiment 5.
Hereinafter, the fifth embodiment of the present application will be described with reference to the parts different from the first embodiment.
When an abnormality occurs in the treatment system of a water and sewage treatment facility, the cause must be identified and operations must be performed to return the treatment facility to a normal state.
The state determination device 500 of the fifth embodiment is proposed as a solution to such a problem, and will be described below with reference to FIG.

FIG. 6 is a block diagram showing a schematic configuration of the state determination device 500 according to the fifth embodiment.
The state determination device 500 of the present embodiment adds a corresponding operation extraction unit 578 and a corresponding operation information database 579 (hereinafter, referred to as a corresponding operation information DB) to the state determination device 100 shown in the first embodiment. It is a thing.
Here, the corresponding operation extraction unit 578 records the operation performed by the operator on the water and sewage treatment facility within the set period after the state determination unit 75 determines that the condition is abnormal, and then extracts the operation. And present it to the operator. The details will be described below.

The response operation extraction unit 578 records the control operation by the operator within a certain period of time after the state determination unit 75 detects that an abnormality has occurred in the processing system in the determination control in the response operation information DB 579 as operation pattern information.
Then, the corresponding operation extraction unit 578 excludes operation patterns of operations having a significantly different type of processing system in which an abnormality has occurred and operation patterns of different control types from the operation patterns recorded in the past in the corresponding operation information DB 579. Above, a similar operation pattern is extracted. Then, the frequency of this similar operation pattern is calculated and registered in the corresponding operation information DB 579. Here, the above-mentioned operations having different control types include, for example, an operation for water quality control of water treatment and an operation for electrical equipment control.

Here, the operation pattern information refers to a control setting value set by an operator, an operation order such as an input order of setting values when a plurality of controls are executed, and an operation amount.
Further, when the operation related to the same abnormality is registered in the corresponding operation information DB 579 in advance, the corresponding operation extraction unit 578 performs statistical processing and standardizes the operation pattern.

Then, when a new abnormality occurs in the processing system, the corresponding operation extraction unit 578 sets this operation pattern when the operation pattern information related to the abnormality determined by the state determination unit 75 to be abnormal is recorded in the corresponding operation information DB 579. The information is extracted and displayed to the operator as reference information.

The state determination device of the present embodiment configured as described above is
In the determination control, the determination unit
When an abnormal processing system, which is the processing system determined to be in an abnormal state, is detected, the operation by the operator for the processing system during the set period from the time when the abnormality is detected is recorded as operation pattern information.
When the abnormal processing system is newly detected after the lapse of the set period, the recorded operation pattern information related to the type of the processing system of the detected abnormal processing system is extracted and output.
It is a thing.

In this way, the determination unit records the operation by the operator within the set period as the operation pattern information from the time when the abnormal state is determined. Then, when the abnormal processing system is newly detected, the operation pattern information related to the type of the processing system of the newly detected abnormal processing system is extracted and presented to the operator. In this way, by automatically extracting the response operation at the time of abnormality from the past operator operation history, it is possible to reduce the design load and quickly respond to the abnormal state.

Although the present application describes various exemplary embodiments and examples, the various features, embodiments, and functions described in one or more embodiments are applications of a particular embodiment. It is not limited to, but can be applied to embodiments alone or in various combinations.
Therefore, innumerable variations not illustrated are envisioned within the scope of the techniques disclosed in the present application. For example, it is assumed that at least one component is modified, added or omitted, and further, at least one component is extracted and combined with the components of other embodiments.

10,21,22,23,24,30,41,42,43,44,50 processing system, 60 processing network, 70 judgment unit, S sensor (detection unit), 20,40 parallel processing system, 100,300, 400,500 Status judgment device.

Claims

In a processing network having a plurality of processing systems for processing an object to be processed
A detection unit that detects the state of each of the processing systems as measured value information,
A determination unit that performs determination control for determining the state of the processing system based on the detected measured value information is provided.
In the determination control, the determination unit
A reference determination formula for determining the state of the processing system based on the correlation between the measured value information of each processing system is used.
Status judgment device.
The determination unit
It has processing process information indicating the connection relationship of a plurality of the processing systems in the processing network, and has processing process information.
In the determination control, a combination of a plurality of the processing systems is set based on the connection relationship of the processing systems shown in the processing process information, and the reference commonly used for each of the set combinations. Create a judgment formula,
The state determination device according to claim 1.
The processing network is configured to have a parallel processing system that processes the object to be processed in parallel by the plurality of processing systems.
In the determination control, the determination unit
Based on the connection relationship shown in the processing process information, the reference determination formula commonly used for each combination of the processing systems having a parallel relationship in the parallel processing system is created.
The state determination device according to claim 2.
The processing on the object to be processed is carried out stepwise by the plurality of the processing systems.
The processing process information further indicates the processing order of the processing system to be performed stepwise.
In the determination control, the determination unit
When an abnormal processing system, which is the processing system determined to be in an abnormal state, is detected, the processing belonging to the front stage side or the rear stage side of the abnormal processing system is based on the processing order shown in the processing process information. Judging the system status,
The state determination device according to claim 2 or 3.
In the determination control, the determination unit
The first distribution information of a plurality of the measured value information detected in the first period including the period of the abnormal state in the abnormality handling system, and the detection in the second period prior to the first period. At least one of the second distribution information of the measured value information of the plurality of the abnormal processing systems, and
In each of the processing systems other than the abnormality processing system, the third distribution information of the plurality of the measured value information detected within the set period is derived.
Using at least one of the first distribution information and the second distribution information and the third distribution information, the state of each of the processing systems other than the abnormal processing system is determined.
The state determination device according to claim 4.
The determination unit
It has a plurality of statistical processing formulas for deriving an evaluation value based on a plurality of the measured value information detected within a set period.
In the determination control, the evaluation values corresponding to the statistical processing formulas are derived for each of the first period and the second period.
The statistical processing formula having the largest difference in the evaluation value between the first period and the second period is selected from a plurality of the statistical processing formulas, and the selected statistical processing formula is used as a new reference determination formula. Used as
The state determination device according to claim 5.
The statistical processing formula is
The evaluation value is configured to indicate at least one of a maximum value, a minimum value, an average value, and a rate of change of a plurality of the measured value information detected within a set period.
The state determination device according to claim 6.
The processing network is configured to have a parallel processing system that processes the object to be processed in parallel by the plurality of processing systems.
The determination unit
The processing of the third distribution information of the measured value information as the input value of the reference determination formula in the determination control and the fourth distribution information of the calculation result as the calculation value of the reference determination formula in the parallel processing system. Derived for each combination of systems,
Correction to correct the reference discriminant so that the determination result in the determination control is the same in each processing system in the parallel processing system based on the derived third distribution information and the fourth distribution information. Create a coefficient,
The state determination device according to any one of claims 2 to 7.
In the determination control, the determination unit
When an abnormal processing system, which is the processing system determined to be in an abnormal state, is detected, the operation by the operator for the processing system during the set period from the time when the abnormality is detected is recorded as operation pattern information.
When the abnormal processing system is newly detected after the lapse of the set period, the recorded operation pattern information related to the type of the processing system of the detected abnormal processing system is extracted and output.
The state determination device according to any one of claims 2 to 8.