JP3173688B2 - Plant monitoring equipment - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a plant monitoring device for monitoring a plant.

[0002]

2. Description of the Related Art Process data is monitored by checking an upper limit, a lower limit, a rate of change, and a deviation from a target value of a data value. This method is effective for abnormal changes in data, such as abnormal reactions, abnormal behaviors, or sensor failures, and has been widely applied. But,
The upper / lower limit check for the change data at a minute level over a long period of time is already too late when the alarm is generated, and is not necessarily effective. Also,
In the alarm output, the relationship with other instruments cannot be described in the alarm processing for each instrument, and if the status of the entire process is not grasped, the essence of the cause of the abnormality is lost due to the chain of alarms. , Which places a heavy burden on the operator.

[0003]

SUMMARY OF THE INVENTION It is an object of the present invention to provide a plant monitoring apparatus which reduces the burden on operators.

[0004]

SUMMARY OF THE INVENTION The present invention relates to a plant monitoring apparatus for monitoring a plant.
Of the surveillance area in advance and focus sequentially on the divided area
The target range determining means to be focused on and the target range determining means
When the point is scored, the current plant condition
Infers the point of interest depending on whether or not
Attention part determination means determined by
Data recognition means for recognizing the state of the process data at the point of gaze of the means , inference from the state of the process data recognized by the data recognition means, and abnormality prediction means for predicting an abnormality; And avoidance operation means for performing an avoidance operation on the other hand.

[0005]

According to the present invention, attention range determining means, note
The eye portion determination means determines which portion of the plant is to be the point of interest. Then, the state of the process data of the gazing point is recognized by the data recognition means. The abnormality predicting unit predicts an abnormality based on the state of the process data. Based on the result of the abnormality prediction, the avoidance operation means operates the plant to avoid the plant from the abnormality.

[0006]

BRIEF DESCRIPTION OF THE DRAWINGS FIG. FIG. 1 is a configuration diagram showing one embodiment of the present invention. In the figure,
Reference numeral 1 denotes a plant, and reference numeral 2 denotes a gazing point determining unit that determines which part of which range of the plant 1 is to be the gazing point. Numeral 3 is a data recognizing means for recognizing the state of the process data of the gazing point of the gazing point determining means. Reference numeral 4 denotes an abnormality predicting unit that infers from the state of the process data recognized by the data recognizing unit 3 to predict an abnormality. Reference numeral 5 denotes an avoidance operation means for performing an avoidance operation on the plant 1.
The avoidance operation means 5 includes a manual operation means for manually operating the plant 1, a rule-based inference means for inferring an abnormality avoidance operation by a rule and operating the plant 1, and a past case. And a case-based inference means for inferring anomaly avoidance based on the operation of the plant 1.

In the gazing point determining means 2, reference numeral 21 denotes an attention area determining means for determining which area of the plant 1 is to be noted. Reference numeral 22 denotes attention portion determination means which extracts a condition from the condition portion 23 and determines which portion is to be the point of interest. Then, the condition unit 23 pays attention to a default rule that basically determines which part of the range is to be the point of gaze and, for example, what kind of driving is being performed and what kind of product is being made. The range stores the condition of the point of gaze determined by what function it has. In the abnormality prediction means 4, reference numeral 41 denotes inference means for inferring whether or not the plant 1 will become abnormal in the future by using a model 42 relating to the plant 1 (for example, a causal relation model between parameters that determines the behavior of the target).

The specific structure of the avoidance operation means 5 is shown in FIGS.
It is shown in FIG. FIG. 2 is a block diagram of the manual operation means.
Is a block diagram of the rule-based inference means, and FIG. 4 is a block diagram of the case-based inference means. In FIG. 2, reference numeral 51a denotes a display unit which displays the state of the plant, the reason for predicting the abnormality, and the like in response to the abnormality prediction by the abnormality prediction unit. Reference numeral 51b denotes an operation unit that performs an operation on the plant in response to an operation by an operator. Reference numeral 51c denotes a case storage unit which stores the contents operated by the operation unit 51b and the contents of the abnormality in the case base 51d.

In FIG. 3, reference numeral 52a denotes inference means for extracting a rule from a rule base 52b for storing a rule of an avoidance operation for a certain abnormality and inferring an avoidance operation. Reference numeral 52c denotes an operation means for operating the plant based on an operation command from the inference means 52a.

In FIG. 4, reference numeral 53a denotes an abnormal feature extracting means which receives the contents of the abnormalities of the plant from the abnormal predicting means 4 and extracts abnormal characteristics. A search unit 53b searches the case base 51c for case data most similar to the abnormal feature. An application unit 53c receives case data from the search unit 53b, and causes the simulation unit 53d to perform a simulation based on the model 53e using a past operation procedure described in the case. Then, it is determined whether the operation is normal. If it operates normally, perform an avoidance operation on the plant. If it does not operate normally, the case data and the simulation data are sent to the correction means 53f. The correction unit 53f corrects the operation data in the case data by a certain rule or manually to make the operation procedure match the current abnormal content, and performs the avoidance operation on the plant. And 53g
Is a storage unit that stores, in the case base 51c, the contents of operations and abnormalities operated by the application unit 53c and the correction unit 53f as new cases.

The operation of such a device is described below.
FIG. 5 is a flowchart showing the operation of the apparatus shown in FIG. The gazing point of the plant 1 is determined by the gazing point determining means 2. Then, the data recognizing means 3 recognizes what state the process data of the gazing point of the plant 1 is in.
Based on the recognition result, the abnormality prediction unit 4 predicts an abnormality of the plant 1. If there is no possibility of abnormality, the operation returns to the initial operation, and the gazing point determining means 2 operates. If there is an abnormality, the avoidance operation means 5 performs an abnormality avoidance operation on the plant 1. Then, the gazing point determining means 2 operates again.

The above operation will be described with a specific example. FIG. 6 is a diagram showing a specific configuration of the plant 1, FIG. 7 is a diagram showing process data of gazing points c and e in a range b of the plant 1, and FIG.
FIG. 7 is a diagram for explaining an operation of predicting an abnormality from the process data of the gazing points C and E using the model in the range b. The gazing point determining means 2 selects the range b from the ranges a to e, and determines the gazing point as the gazing points C and E. The state recognizing means 3 extracts features from the process data of the gazing points c and e, and recognizes what state it is at a certain time t. Then, the abnormality prediction unit 4 predicts an abnormality using the model in the range b. Here, the division of the range is constituted by a unit or a unit group. A unit is a logical device composed of a functionally significant device group. For example, a distillation apparatus, which is a combination of apparatuses, is called a unit.

Next, the specific operation of each configuration will be described. 9 and 10 are flowcharts showing the operation of the gazing point determining means 2. FIG. 9 shows the operation of the range-of-interest determination means 21, and FIG. The range-of-interest determination means 21 focuses on the divided range of the plant 1 determined by the definition. Then, after a lapse of a predetermined time, the focus is on the range of the next plant. This operation is repeated. Here, the definition is a repetitive operation,
There are temporal operations, jump operations, and skip operations.

The part-of-interest determination means 22 determines which condition in the condition section 23 the current state of the plant 1 satisfies for the range of the plant 1 focused by the range-of-interest determination means 21. Judge from process data. If none of the conditions is satisfied, a feature point is determined using a default rule. If the condition is satisfied, the gazing point is determined based on the condition. This process is repeated during a period when one area of the plant 1 is focused.

FIG. 11 is a flowchart showing the operation of the data recognizing means 3. The data recognizing means 3 extracts time-series data (for example, trend data) of the point of regard. Then, the time series data is recognized as a pattern. For example, it is recognized as a pattern such as increase, decrease, and equilibrium. Save this recognized result. The next time series data is extracted and the same operation is performed for the number of gazing points. When the number of gazing points has been reached, the data is sent to the abnormality prediction means 4.

FIG. 12 is a flow chart showing the operation of the abnormality prediction means 4. The inference means 41 infers an abnormality predicted based on the model 42, and sends information of the abnormality to the avoidance operation means 5 if there is a possibility of abnormality. If there is no possibility of abnormality, do nothing.

Operation of Avoidance Operation Means 5 FIG. 13 is a flowchart showing the operation of the avoidance operation means 5. Whether to perform the manual operation, the rule-based inference, or the case-based inference is determined in advance, and the avoidance operation is performed by means determined as follows. (1) Manual operation means The current state of the plant and the reason for predicting the abnormality are displayed on the display means 51a. The operator performs an avoidance operation on the plant with reference to this. Then, the case storage means 5
1c stores the operation contents and the abnormal contents from the operation means 51b as a case in a case base.

(2) Rule-based inference means The inference means 51a extracts a predetermined rule from the rule base 52b and infers an operation to avoid an abnormality. In response to the inference of this operation, the operation means 52c performs an avoidance operation on the plant.

(3) Case-based reasoning means The abnormal feature extracting means 53a receives the contents of the plant abnormality from the abnormal predicting means 4 and extracts the abnormal features. Based on the feature of this abnormality, the search means 53b searches the case base 51c for the most similar case data. Then, based on the case data, whether or not the application unit 53c is applicable to the plant, the simulation unit 53d is operated using the past operation procedure described in the case. The simulation means 53d performs a simulation using the model 53e, and determines whether the plant operates safely. When it does not operate safely, the case data and the simulation data are sent to the correction means 53f. Correction means 53
f receives the case data and the simulation data, modifies the operation for the plant, and sets an operation procedure that matches the current abnormal content. And the storage means 53g
Stores the contents operated by the application unit 53c and the correction unit 53f and the abnormal contents as a new case in the case base 51c. Finally, the application unit 53c or the correction unit 53f performs an avoidance operation on the plant.

Since the operation is performed as described above, the monitoring is automatically performed, and the burden on the operator is reduced.

[0021]

According to the present invention, the range of interest determination means,
The gaze point is determined by the eye portion determining means , the state of the process data is recognized by the data recognizing means, and the abnormality is predicted by the abnormality predicting means from this state. Avoid by means. As a result, the monitoring is automatically performed, so that there is an effect that the burden on the operator is reduced.

[Brief description of the drawings]

FIG. 1 is a configuration diagram showing one embodiment of the present invention.

FIG. 2 is a configuration diagram of a manual operation unit of the avoidance operation unit 5;

FIG. 3 is a configuration diagram of a rule inference unit of the avoidance operation unit 5;

4 is a configuration diagram of a case-based inference unit of the avoidance operation unit 5. FIG.

FIG. 5 is a flowchart showing the operation of the apparatus of FIG. 1;

FIG. 6 is a specific configuration diagram of a plant 1.

FIG. 7 is a diagram showing process data of gazing points c and e in a range b of the plant 1;

FIG. 8 is a diagram for explaining an operation of predicting an abnormality using a model in a range b from process data of gazing points C and E;

FIG. 9 is a flowchart showing the operation of the range-of-interest determination means 21;

FIG. 10 is a flowchart showing an operation of a focused part determining means 22;

FIG. 11 is a flowchart showing the operation of the data recognition means 3;

FIG. 12 is a flowchart showing the operation of the abnormality prediction means 4;

FIG. 13 is a flowchart showing the operation of the avoidance operation means 5;

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Plant 2 Point-of-interest determination means 3 Data recognition means 4 Abnormality measurement means 5 Avoidance operation means 21 Focus area determination means 22 Focus part determination means

Continuation of the front page (56) References JP-A-64-82205 (JP, A) JP-A-4-199202 (JP, A) JP-A-59-117608 (JP, A) JP-A-1-2655311 (JP) , A) (58) Field surveyed (Int. Cl. ⁷ , DB name) G05B 23/02 G05B 13/02

Claims (1)

(57) [Claims] In a plant monitoring apparatus for monitoring a plant, a monitoring range of a monitoring target plant is divided in advance, and the divided
A focus range determining means for sequentially focusing on the range , and a current plant when the focus range determining means focuses on the current plant.
Depending on which condition the condition of
Attention part determination means that determines by inference whether to make a part
And a data recognizing means for recognizing the state of the process data at the point of gaze of the remarkable part determining means , and an abnormality prediction for inferring from the state of the process data recognized by the data recognizing means to predict an abnormality And a avoidance operation means for performing an avoidance operation on the plant.