JPH0363082B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION When a failure occurs in one of the component equipment in a plant, the present invention predicts to which range of equipment in the plant the failure will spread after a certain period of time,
The present invention relates to a plant failure influence range prediction/display device that displays the predicted influence state on a network showing the coupling relationship of plant component equipment. This device can be used to predict and display the scope of failures in nuclear power plants, water and sewage systems, chemical plants, etc.

Conventionally, in order to predict the failure spread range, (a)
Fault Tree Analysis (abbreviated as FTA) and Event
Methods such as Tree Analysis (abbreviated as ETA) that recognize the cause-and-effect relationships of failures on tree diagrams, (b) manual simulation methods that use blueprints, and (c) hardware staining methods that use actual equipment or model equipment. Uration method has been mainly used.

However, the recognition method (a) above focuses on qualitatively capturing the cause-and-effect relationship of failures on a tree diagram.
It is difficult to quantitatively understand the scope of failure and its changes over time. In addition, the manual simulation method (b) above is a method that predicts the range of failure spread after t hours by manually following the design drawings. The propagation path becomes complex, making it difficult to completely know the extent of the failure's propagation. The hardware simulation method (c) uses an actual device or model device with a scale reduced to 1/20, which not only makes it difficult to manufacture the device, but also involves danger when actually performing the simulation.

As a means of displaying the failure influence range, (a) above displays a tree diagram, so the same device appears in multiple locations, making it difficult to understand the failure influence range from the coupling relationships of system component devices. In (b) and (c) above, all the propagation relationships of failures, including changes over time, must be stored in the storage device in advance, which has the drawback of requiring a huge amount of time and storage capacity to determine the propagation relationships. be.

The present invention has been made in order to solve the above-mentioned drawbacks of the prior art, and provides a device that can easily predict the failure influence range and display it on the network in order to make it easier to visually understand the failure influence prediction range. This is what we provide.

Hereinafter, the present invention will be explained in detail with reference to Examples.

In FIG. 1, a detector 10 is attached to each component of a plant 101 consisting of a plurality of components.
2 is provided. These detectors 102 detect the operating state of each component, such as flow rate, temperature, frequency, etc., and output a detection signal 106 to the failure diagnosis device 103. The failure diagnosis device 103 includes:
A detection signal obtained from each component device when each component device is in a normal state (hereinafter referred to as a normal signal) is stored in advance, and this normal signal and the detection signal 106 are compared and the difference is determined as a predetermined value. The above equipment, that is, the faulty equipment is detected, and the faulty equipment number signal 107 is output to the fault spread prediction/display control device 140. A specified time signal 109 representing the time from the current time to the predicted time is input to the device 140 by the input device 110 .
As will be described in detail later, the device 140 predicts the range in which the condition will spread within the time specified by the specified time signal 109 based on the failed equipment number signal 107, and creates a network indicating the coupling relationship of the plant component equipment. Display device 1 displays the predicted spread range along with the workpiece.
Display as 05.

FIG. 2 is a block diagram of a failure spread prediction/display control device. In the memory 143, in advance,
A fault wave related matrix A is stored that corresponds to the presence or absence of direct fault propagation between component devices e _i and e _j of the plant. This matrix A is a matrix as shown in the transition table 201 in FIG. 3, and the element a _ij in the i row and j column of the matrix A is
This is the direct failure propagation time from equipment e _i to e _j , and is a value determined based on actual simulation results. For example, a _1o is 100, indicating that the failure propagation time from device e ₁ to device e _o is 100 seconds. Note that if a _ij = 0, the device e _i to device e _j
This indicates that the failure does not spread to other areas.

The transition matrix arithmetic unit 145 reads the transition table 201 from the memory 143 and reads out the transition table 201 from the memory 143, excluding devices that do not cause a failure to spread to other devices and to which other devices do not have a failure, including direct and indirect spillovers. A transition table 206 representing the failure propagation time between devices is calculated and outputted to the memory 144. This calculation is performed as follows. Elements of matrix A of transition table 201
Convert all values of a _ij ≠ 0 to 1 and create the transition table 202
The matrix B of (element is b _ij ) is calculated (transition table 202
is output to the display control device 146. ). For example, a _1o of matrix A is 100, but this is converted to 1, and b _1o of matrix B is set to 1. Furthermore, an identity matrix is added to matrix B to calculate transition matrix C shown in Table 203. This matrix C is changed until C ^m =C ^m+1 (m+
1) Multiply to calculate the transition matrix D shown in the transition table 204. Among the elements d _ij of the matrix D, i is determined in which the elements in the i-th row and the i-column are all 0 except for d _ii . For example, in the transition table 204, this corresponds to elements in 10 rows and 10 columns. Therefore, from transition table 201, 10 rows 10
After removing the column, a transition matrix F (element f _pq ) shown in the reduced transition table 205 is obtained.

Next, transition matrix F (element f _pq ) shown in transition table 205
From this, the transition matrix G (element g _pq ) shown in the transition table 206 is obtained. For this purpose, the following equations (1) to (4) are calculated.

g _pq = {f _pq (1), f _pq (2), ..., f _pq (L)} ...(1) f _pq (1) = f _pq ...(2) f _pq (k) = - min r=1;2...,n{f _pq (k-1)f _rq ...(3) f _pr (k-1)f _rq = f _pr (k-1)+f _rq (f _pr (k-1)・If f _rp ≠ 0) 0 (f _pr (k-1) ・If f _rp = 0) ...(4) Here, equation (1) is f _pq (1), f _pq (2) ,...f _pq (L),
This is an operation to find the minimum value excluding 0. However, when all elements f _pq (i) are zero, it is zero.
L is a predetermined positive integer, and is given as 3, for example. f _pq (i) indicates the time when a failure spreads through (i-1) devices. Therefore, considering the special case of i=1, f _pq (1) indicates the time when the failure directly spreads. (3)
The formulas are f _p1 (k-1) f _1q , f _p2 (k-1) f _2q ,...
This is an operation to find the minimum value excluding 0 out of f _po (k-1) f _oq . However, all elements f _pr (k−
1) It is zero when f _rq is all zero.

Therefore, equation (1) means that when a failure propagates from p to q, a propagation route with the minimum propagation time and the number of devices passing through on the way is searched.

The transition table 206 obtained by the transition matrix calculation device 145 in this manner is stored in the memory 144.

On the other hand, the memory 142 stores a transition probability table 207 shown in FIG. Although only 1.0 is illustrated as the value of the transition probability in FIG. 4, the value of the transition probability is generally any real value between 0.0 and 1.0.

The failure propagation prediction device 141 stores in the memory 144 a time signal for the failure to spread from the failed device e _i specified by the failed device number signal 107 to other devices e _j .
Read from the transition table 206 stored in . It is determined whether the read time signal _T1 is greater than zero and smaller than the designated time signal 109, and if this is true, the designated time signal 109 is determined.
Since it is predicted that the failure will spread within the time specified by , the corresponding device e _k1 is estimated to be the failure spreading device. Further, a time signal T 1 at which a failure spreads from device e _k1 to other devices e _j is read from the memory 144, and a time signal T ₁ at which a failure spreads from device e _i to e _k1 is read out from the memory 144.
is added to this to obtain the time signal _T2 . A device e _k2 for which the time signal T ₂ obtained in this way is larger than the time signal T ₁ and smaller than the designated time signal 109 is found. Do something like this with the equipment
Repeat m times until e _kn cannot be found. Then, the devices obtained as devices e _k1 , e _k2 , . . . e _kn are outputted to the display control device 146 as failure propagation devices e _k .

The failure propagation prediction device 141 determines from the memory 142 the devices whose failure is predicted to spread from the failed device e _i within the time corresponding to the specified time signal 109.
The failure transition probability P(e _i , e _k1 ) to e _k1 is read. The transition probability P(e _i , ekm) of equipment e _kn is P(e _i , e _kn )=P(e _i , e _k(n-1) ) ・P(e _k(n-1) , e _kn ) ...Determined by calculation (5). The transition probability obtained in this way is determined by the failure propagation equipment e _k in the display control device 146.
is output with

The display control device 146 displays a failure ripple relationship diagram of the power supply 1, pump 2, motor 3, solenoid valve 4, vacuum pump 5, and speed reducer 6 corresponding to each component of the plant shown in FIG. 5a from the transition table 202. , well-known
It is determined by the Interpretive Structural Modeling method and output to the display device 105. A group of devices to which a failure propagates and a network formed by connecting these devices with line segments and representing the failure propagation relationship between the devices are displayed on the display device 105 as shown in FIG. 5B. Further, the display control device 146 controls the display device 105 to display the failed device e ₄ in red based on the failed device number signal 107. Also,
Based on the output of the failure propagation prediction device 141, the failure propagation line from failure propagation device e ₆ and device e ₄ to e ₆ is displayed in yellow, and the failure propagation probability of 1.0 is also displayed. The device 105 is controlled to display the diagram shown in FIG. 5c on the display device 105. Note that the transition direction may also be displayed.

As described above, according to the present invention, it is possible to calculate the predicted range of failure influence after a specified time and the probability of failure influence. Furthermore, by displaying these results on the network, the predicted range of failure spread can be grasped in an easy-to-understand manner, making it easier to take measures against failures.

[Brief explanation of drawings]

1 to 5 are explanatory diagrams of the present invention.

Claims (1)

[Claims] 1. Provided in each component of the plant,
A plurality of detectors detecting the operating status of each component device, and are coupled to the plurality of detectors, and store normal signals obtained from the plurality of detectors when each component device is operating normally. Then, compare this normal signal with the detection signals of the plurality of detectors,
A fault diagnosis device that detects faulty equipment, an input device that inputs a specified time from the current moment to the time when the failure is expected to spread, and information representing the operating status of the plant and the range of fault spread up to the specified time for each component. a display device for displaying in addition on a network representing the connection relationship between The control device stores in advance a first storage device that stores the direct propagation time of a failure between each component device, and a first storage device that stores the direct propagation time of a failure between each component device, including direct propagation and indirect propagation. a transition matrix calculation device that calculates from pre-stored direct propagation times; a second storage device that stores failure propagation times that are elements of the determined transition matrix; and failure propagation times from the faulty device to other devices. is predicted from the failure propagation time between each component device stored in the second storage device, the predicted failure propagation time is compared with the specified time, and the component corresponding to the specified time or less is selected. , a failure propagation prediction device that detects the component equipment within the range where the fault spreads within the specified time, the component equipment that is within the range where the detected fault spreads, and the component equipment within the range where the detected fault spreads. Prediction of fault spread range of a plant, characterized in that the fault spread line is formed on a network representing the coupling relationship between each component, and a display control device generates a display control signal to be displayed on the display device.・Display device. 2. Provided for each component of the plant,
a plurality of detectors that detect the operating status of each component; , a fault diagnosis device that compares this normal signal with the detection signals of the plurality of detectors and detects a faulty device; an input device that inputs a specified time from the current moment to the time when the fault spread should be predicted; a display device that displays information representing the operating status and the range of failure spread up to the specified time in addition to the network representing the connection relationship between each component, and a display device that is connected to the fault diagnosis device, input device, and display device The failure propagation prediction/display control device includes a first storage device that stores the direct propagation time of a failure between each component device, and a failure propagation time between each component device. a transition matrix calculating device that calculates a ripple time including direct ripples and indirect ripples from the direct ripple times stored in advance in the first storage device;
a second storage device that stores failure propagation times that are elements of the determined transition matrix; a third storage device that stores failure transition probabilities between each component device; and a third storage device that stores failure propagation times that are elements of the determined transition matrix; A configuration in which the propagation time is predicted from the failure propagation time between each component device stored in the second storage device, the predicted failure propagation time is compared with the specified time, and the configuration corresponds to the failure propagation time that is less than or equal to the specified time. The device is detected as a component device within the range where the fault spreads within the specified time, and the probability of failure transition to the component device within the range where the detected fault spreads is determined for each device stored in the third storage device. A failure propagation prediction device that calculates from failure transition probabilities between component devices, component devices in the range where a detected fault spreads, failure transition probabilities to those component devices, and configurations in the range where the detected fault spreads. A fault spread range of a plant characterized by comprising a fault spread line to equipment on a network representing a coupling relationship between each component equipment, and a display control device that generates a display control signal to be displayed on the display device. Prediction/display device.