JPS60205707A - Plant monitor device - Google Patents

Abstract

PURPOSE:To grasp surely even abnormal events which occur transitionally in the plant operation mode, by providing a circuit which holds abnormality detection inputs from a plant. CONSTITUTION:Signals outputted from detectors 3 and 7 of a device 2 are converted by corresponding converters 5 and are inputted to a limit check 10 of a computer 8. When a signal 11 is outputted from the check 10, a plant diagnostic logic 9 is started by a start instruction 12, and the signal 11 is detected by a node 13. An abnormal event holding circuit 30 which issues an output 29 holding a logical signal is provided in the output side of a redundant processing part 28, and even transitional abnormal events are transmitted to the plant diagnostic logic 9. If abnormality can be avoided thereafter, a holding release input 31 is inputted by the operation of an operator or the like.

Description

DETAILED DESCRIPTION OF THE INVENTION [Technical Field of the Invention] The present invention relates to a conventional plant monitoring device 1 that monitors a plant and outputs an appropriate message when an abnormality is detected.

[Conventional technology]

Conventionally, there has been a device as shown in No. 11 as this type of device. In FIG. 1, (1) is a plant, (2) is a device that constitutes a part of the plant (1), and has the characteristic that if it becomes abnormal, the influence will spread to other parts of the plant (1), (3) refers to equipment (some kind of abnormality that occurs)
A detector extracts the signal (e.g. high temperature, low pressure, etc.)
4) Output. (5) is a converter that converts the voltage level of signal (4), and outputs signal (6). 1 (7) is a detector that extracts process breath that the process detected by detector (3) has some causal relationship with; (8) is a computer that includes plant diagnosis logic (9) to be described later and a display (not shown) It has a device such as a CRT (CRT).

(0) is the plant diagnosis logic provided in the computer (8), αQ inputs the signal (6) of the converter (5),
A signal (binary signal @) is generated by a limit check using computer application software that outputs an earlier level than the normal alarm level of the plant (1), that is, a level on the safe side of the plant (1). What to output. (b) is a start command (signal) for the plant diagnosis logic (9).

In the plant diagnosis logic (9), when the limit check Ql signal (B) turns on, an observable point (hereinafter referred to as a node) is detected as a warning in the plant diagnosis logic (9). ), α◆ is a node detected in the plant diagnosis logic (9) in the same way as a node, and is a node (observable information) connected to the extractor (7) of the plant (1), and a[ is a node AND gate with 0 as input, QG is OR gate, 071 is the minimum time delay τ3 among the time delays that occur when going from lower node 0 to upper node 01, and trowel is a warning when node (to) is detected and the cause is determined to be in node 041, and is outputted, and is a message consisting of an instruction to prevent the abnormal state from spreading to others. aI is the logic of time delay τ3 and τ4 from node O. The node detected after summation, the trout is the node in the final state where the abnormality extends, 2υ
is a point, and if it becomes abnormal, it will also affect the detector (3), but it is a virtual point that cannot be detected or is not provided with a detector. 4 is a false logic and an unobservable point C! This logic is set up in response to υ and simulates the path through which its influence spreads. (E) is a virtual node corresponding to point (C). (Wealth) is point guυ, temporary empty logic (Wealth),
This is a message that is additionally provided as a node so as to correspond to the node (to) 1ζ, and instructs to avoid the abnormal state.

), 01 is the device (detector in 21, detector (3)
, to).

This is the point where the same parameters are generated in ().

Therefore, the detector (3, (b), □□□ is the point (b)
It is also the same one set up in the Masu takes in the master power from Limit Check On, for example, by majority vote?
The redundant processing unit that selects the redundancy outputs a signal -.

Next, 1. I will explain about I+ work.

The signals output from the detectors (3) and (7) of the device (2) are converted by corresponding converters (5) and input to the limit check Q1 of the computer (8). When the signal (lη) is output from the limit check 01, that is, when it turns on, the plant diagnostic logic (9) is activated by the activation command a2, and the signal Oυ is detected at the node a3. , it is known in advance that the influence is from the detector (7) or others,
The logic of the causal relationship is the AND gate oFj and the OR gate.
It is simulated by gate oar. Therefore, even if the plant diagnostic logic (9) is activated, at that point,
If an abnormality is detected before the node (+41) reaches the W level, the cause is that the first event F occurred when the node (141) was detected.

Furthermore, the cause of the abnormality in detection g (3) is
It is determined that there is an abnormality at J, and a message □□□ is output to Iζ operation p in order to prevent this Iξ abnormality from spreading.

Also, node 01 is detected and the plant diagnostic logic (
9) is tT′ multiplicity 1 “When 1ζ, node (1, 4
When detected, the logic above it is examined. As a result, the next possible state of the plant (υ) is predicted,
The operator can take precautions to prevent this from reaching node (2) in the final state W-. In addition, node 03
Since a conceivable minimum clear distance τ3 is set between node O0 and node O0 in plant (1), node 0:
If node Ql is detected within a shorter time than time τ3 after I is detected, it is determined that the detector (2) on the plant (1) side that detects node a1 has rarely failed. do.

In order to convey these causes and future predictions to the operating month, this plant diagnosis logic (9) is displayed on a display device in the format shown.
The awards and messages of detectors (3) and (7) are additionally displayed on the displayed nodes 03 and 041, respectively. This allows the operator JJ1 to accurately grasp the state of the plant (1) and accurately perform operations to avoid abnormal conditions.

However, after an abnormality is detected at node 03, in order to investigate the cause of the abnormality, nodes below node (2) are investigated, but no abnormality is detected in any node (cause identification failure). determines that the cause of the abnormality is an abnormality at an unobservable point e within the plan l-11), that is, the node (to) is abnormal.

It should be noted that this message contains the content that the cause of the abnormality is in an unobserved node, for example, node (c), and also includes appropriate operation instructions in accordance with this.

In addition, in the device (2), a detector (3
), (2), and (e) are provided multiple times, and a large number of the same parameters are extracted. Similarly, converter (5) and limit check aQ1 are multiplexed, and the 4a 5. Input L to the redundancy processing circuit (2).

Therefore, for example, even if some of the detection paths (3), 2) and 2) are -8ti+ or erroneously created, the redundancy processing section (c)'s part 9 is correct. Treated as 4'
1. Become a believer/3. Plant using such signal Qt% as human power 1. :[Evening 11th [Logic (0) is,
Do not turn on or off nodes 9.1J, 9, etc. carelessly, and take care of any abnormalities in the plant (1).
C. Give information.

The number of triplexed detectors is 21! In this case, r, <, yt:, the redundant processing unit eae may be provided on the plant side in order to reduce the number of cables between the plant (1) and the computer (8). good.

Conventional plant monitoring equipment has the following configuration, and once an abnormality occurs, once the plant passes a certain mode, the normal event disappears even though there is a cause of the abnormality. This has the disadvantage that all nodes are affected by this, resulting in a misdiagnosis of a plant abnormality and an inaccurate abnormality fi avoidance message.

[Desk of the Invention (, Key 2) This i'+j'Sg was created in order to eliminate the drawbacks of the conventional one as described in 1-1. Is there any difference in setting up a circuit and catching the delayed two-way phenomenon? i:', Instantly catch the timing and evaluate the differences in the plant.
It is an object of the present invention to provide a plant monitoring device that can output reliable messages without interruption.

[Invention glaze example]

Hereinafter, one embodiment of the present invention will be explained with reference to FIG.

In FIG. 2, (1) to (g) represent the same parts as in FIG. ... represents an abnormal event holding circuit, and OfJ represents a holding release signal input.

Next, to explain operation 1, the operations from (1) to even are the same as those of conventional plant monitoring devices.The repeated explanation will be omitted below. An abnormal event circuit that can output output 4 holding logical signals at the exit from the redundant processing section (to) is installed to accurately convey even transient abnormal events to the plant diagnosis logic (9) to prevent erroneous diagnosis. Perform plant monitoring that does not occur. When the abnormality can be avoided, the operator or the like inputs the hold release input op using the same operation as a normal ANN ryat.

Note that although the present invention is implemented using S/W in the description above, it may also be configured using a hard-wired device.

〔Effect of the invention〕

As described above, according to the present invention, even abnormal events that occur transiently during the plant operation mode can be detected reliably, thereby eliminating the possibility of erroneous diagnosis and providing highly reliable information regarding plant abnormalities. Plow the information! (It has the effect of being able to provide transposition.

[Brief explanation of drawings]

1c+4 is a block diagram showing a conventional plant monitoring device, and FIG. 2 is a block diagram showing an embodiment of the plant monitoring device of the present invention. 5 (1)...plant, (3), (7), (g), g →...detector, r5)...converter, (8)...computer, (9)...・Plant diagnosis logic, Q(
1...Limit check, (to), Q4J, O groan,
(4), W...node, light...and gate, (
IG...or gate, aη...delay, @l)...
・Point, (2)...Temporary empty logic, (2)...Redundant processing unit, (2)...Abnormal event holding circuit, 0s...Holding release signal input. Note that the same reference numerals in the figures indicate the same or equivalent parts. Zangetsu/Hito Oiwa Tan class

Claims (1)

[Claims] When the output of a detector that detects a process notification input to a specific part of the plant indicates a second abnormality level that is set on the safer side than a predetermined first abnormality level, the output of the L detector is set to t. In a plant monitoring device that performs the above diagnosis as an input condition of a logical consideration tree for diagnosing the plant, the above detectors are used in multiples, and the outputs of these detectors are selected according to a predetermined logic. What is claimed is: 1. A plant monitoring device characterized by having a redundant processing device for detecting abnormal events, and an abnormal event retaining device.