JPH0444772B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a plant diagnosis device that identifies abnormal events in a plant online in real time in order to contribute to improving the operational reliability and availability of large-scale plants.

A conventional device of this type is shown in FIG. In the figure, 1 is a data acquisition device (for example, an analog-to-digital converter) for reading process data, and 2 is a data collection device that compares the process data with a reference value. arithmetic processing device 1 for converting into “1”;
Storage devices 1 and 4 store cause-and-effect trees (hereinafter referred to as CCT) that describe the propagation sequences of various abnormal events that occur in the plant using logical expressions. The storage devices 2 and 5 identify the first cause of the abnormality using the judgment results stored in the storage devices 1 and 3 and the CCT stored in the storage devices 2 and 4, and display the first cause in the cathode ray tube display device. This is a diagnostic processing device that outputs data to 6.

Next, the operation will be explained.

process data x _i (i=1, 2,...,N)
shall be. x _i is quantized by the data acquisition device 1. Using the quantized data x _i as input, the arithmetic processing unit 2 performs the processing shown in the formula, and as a result, s _i
(i=1, 2, . . . , N) and the event occurrence time are stored in the memory device 3.

_{S i} ⁼ O IF X ^L _i ≦ _{X i ≦} X ^U _i ^S _{i =} ^l _IF The lower limit alarm level, x ^L _i is the upper limit alarm level, and is calculated by the formula.

X ^L _i =X ^R _i −σ _i X ^U _i =X ^R _i +σ _i where σ _i is the variation tolerance and x ^R _i is the process data
This is the reference value of x _i , and is selected and calculated from the following three methods depending on the nature of each process data.

(A) Constant value (B) Function of turbine load, reactor power (C) Separate process data x _j (A) is applied to process data that is controlled to a constant value in the entire load range. (B) is applied to process data whose set values change depending on the load. Second
A typical example is shown in the figure. (C) is applied to process data obtained from multiplexed detectors and output process signals that vary depending on the input process signal to be diagnosed.

The obtained s _i is called the status of x _i . Also,
The time when the status becomes 1 is called the event occurrence time.

The CCT execution processing device is executed by the diagnostic processing device 5 based on information from the storage devices 3 and 4. First disturbance
The following processing is performed to search for the cause.

First, subs to be searched from a large number of CCT groups.
In order to extract the CCT portion, an entry node whose status has changed (designated in advance as a node to start analysis) stored in the storage device 3 is detected, and a search starts from that node to lower nodes. And the first gate?
Find out whether it is an AND gate or an OR gate. Next, it is checked whether the plant observation data written in the storage device 3, which is the state of each input of the gate and the relationship between the event generating devices, matches the CCT model written in the storage device 4. If the logic is established, the next node to be searched is determined according to the CCT model stored in the storage device 4 for expansion to lower nodes. The order of search nodes is to start with the entry node, search for all nodes that are children of the entry node, and then search for all nodes that are grandchildren of the entry node.

In this way, in order, for each gate connected below the entry node, it is determined whether the node on the input side of each gate is false (mismatch) or true (match), and the true branches are spliced together. . When the expansion reaches the primary node, the logic of this entry node is established, and it is determined that the cause has been successfully identified.

If it is found that there is no cut/set for the entry node during the expansion, the diagnosis is terminated at that point.

If the diagnosis is aborted or finished, analysis from the next entry node is performed taking priority into account.

For entry nodes for which the cause has been successfully identified, the logic is traced upwards from that node, and the progress of the disturbance is analyzed. That is,
Check how far the disturbance has spread in the upper direction, and then assume that the disturbance has spread further and check to determine whether the nodes above the topmost active node will change state after some time. Therefore, first check the type of gate immediately above the entry node. It then examines the input status of that gate and its output status is currently
Search to see if it is Falsc but becomes True over time.

As described above, through the cause identification and prediction processing, it is possible to determine a tree that sequentially represents the disturbance from its occurrence to its future propagation.

Restores the message contents and variables connected to this relevant tree. Then, the following information is restored on the cathode ray tube display device 6.

(1) Messages specified in the tree part that has progressed up to now (2) Messages that will become active in the near future (3) Variable list as they are related to the tree part obtained in the previous steps In conventional plant diagnosis devices, logic conversion is performed in the arithmetic processing unit 2 by comparing reference values determined from static characteristics of the plant or input/output process signals of control system diagnosis target elements.
During a plant transition, a temporary status is generated due to delays in the plant's dynamic characteristics and delays in each element, resulting in a drawback such as misdiagnosis.

This invention was made in order to eliminate the above-mentioned drawbacks of the conventional system. Among the signals that enter the processing unit 2, the process data assigned to the primary node is calculated using a dynamic characteristic model of the plant. It is an object of the present invention to provide a plant diagnosis device that does not make a false diagnosis even in all operating modes of the plant by adding a device that performs logic conversion by comparing the plant with a reference value.

An embodiment of the present invention will be described below with reference to the drawings. In FIG. 3, 7 inputs the process data x _i read by the data collection device 1, and compares the process data x _i with a reference value to determine whether the process data x _i is within the allowable range. This is an arithmetic processing device that calculates the reference value using a dynamic characteristic model of each device to be diagnosed in the plant in order to determine the condition.

The process data x _i is quantized by the data acquisition device 1 . Of the quantized data x _i , Figure 2
The process data Xi related to equipment (equipment that is not affected by plant transients, etc.) that can be accurately diagnosed by comparison with the reference value of Process data related to equipment that cannot be accurately diagnosed unless compared to standards calculated using
Xi is input to the arithmetic processing unit 7.

The arithmetic processing unit 7 calculates a reference value using a plant dynamic characteristic model and performs logic conversion. Examples of calculations for each element of the filter, controller, and valve as diagnostic targets are shown below.

(A) Filter The dynamic characteristics of the filter are expressed by the following equation.

y^ _f(k) = y^ _f(k-1) = △t/T _f (U _f(k-1) −y^ _f(k-1) ) Here, △t is the time step, T _f is the time constant,
y^ _f(k) is the output estimate of the filter, U _f is the plant data input to the filter, and k is the time step. If the filter is normal, y^ _f(k)
simulates the filter output signal value y _f(k) .
y^ _f(k) calculated by is the output signal value of the filter
Use it as a reference value for y _f(k) and perform logic conversion using the formula.

X ^L _yf(k) =y^ _f(k) −σ _yf X ^U _yf(k) =y^ _f(k) −σ _yf S _yf(k) =O IF X ^L _yf(k) y _{f(k )} X ^U yf( _{k )} S _yf(k) = l IF y _f(k) ＞X ^U _yf(k) ory _{f (k)} ^< The corresponding status, σ _yf , is the variation tolerance.

(B) Controller (proportional/integral controller) The dynamic characteristics of the controller are expressed by the following formula.

y^ _c(k) =y^ _c(k-1) +k _c (1+△t/T _c )u _c(k) −k _c u _c(k-1) Here, k _c is the proportional gain, T _c is the integral time constant,
y^ _c(k) is the estimated output value of the controller, u _c is the plant data input to the controller, and k is the time step. The calculated y^ _c(k) is used as a reference value for the output signal value y _c(k) of the controller, and is subjected to logic conversion using the formula.

X ^L _yc(k) =y^ _c(k) −σ _yc X ^U _yc(k) =y^ _c(k) −σ _yc S _yc(k) =O IF X ^L _yc(k) y _{c(k )} X ^U _yc ( _{k )} S ^yc _{( k) =} l IF y _c(k) ^> status, σ _yc is the variation tolerance.

(C) Valve The dynamic characteristics of a valve can generally be approximated by a first-order lag element and a nonlinear element.

z^ _v(k) =z^ _v(k-1) +△t／T _v (u _v(k-1) −x^ _v(k-1) ) z^ _v(k) = F(z^ _v(k) ) Here, T _v is the delay time constant, y^ _v(k) is the estimated value of the valve opening, u _v(k) is the signal of the controller that drives the valve, and F is the nonlinear function. . y^ _v(k) calculated by is the valve opening y _v(k)
X ^L _yv(k) = y^ _v(k) −σ _yv X ^U _yv( k) = y^ _v(k) −σ _yv S _yv(k) =O IF X ^L _{yv (} ^k _{) y v (} ^k ₎ ^_ _{_} The status shown is the storage device 3.
is stored in CCT execution processing is performed in the same manner as before.

In the above embodiment, filters, controllers, and valves were explained as calculations by the processing unit 7, but plant main engines (for example, steam generators)
If a dynamic characteristic model, etc., are also included, the status calculations for non-multiplexed detectors can be performed using the same logic.

As described above, according to the present invention, process data read by a data collection device is input,
A calculation process that calculates the standard value using a dynamic characteristic model of each equipment to be diagnosed in the plant in order to determine whether the process data is within the allowable range by comparing the process data with the standard value. Since the equipment was installed, appropriate standard values could be set according to the plant status, and as a result,
It is possible to prevent misdiagnosis during plant transients, and when considering plant transients, it was previously necessary to create a cause-and-effect tree for each plant state, but this method uses a dynamic characteristic model. Since the reference value is calculated based on the system, it is possible to respond to the condition of each plant by simply creating one type of cause-and-effect tree.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing a conventional plant diagnostic device, FIG. 2 is a diagram showing an example of a reference value function of a conventional arithmetic unit 2, and FIG. 3 is a block diagram showing an embodiment of the plant diagnostic device according to the present invention. figure. In the figure, 1 is a data collection device, 2 is a first arithmetic processing device, 3 is a first storage device, 4 is a second storage device, 5 is a diagnostic processing device, 6 is a cathode ray tube display device, and 7 is the present invention. This is a status arithmetic processing device with a dynamic characteristic model added. In addition, one reference numeral in the figure indicates the same or corresponding part.

Claims (1)

[Claims] 1 A data collection device that reads process data is input, and the process data read by the data collection device is input, and it is determined whether the process data is within an allowable range by comparing the process data with a reference value. an arithmetic processing device that calculates the reference value using a dynamic characteristic model of each device to be diagnosed in the plant; a first storage device that stores the determination result of the arithmetic processing device;
a second storage device that stores a cause-and-effect tree that describes the propagation sequence of various abnormal events that occur in the plant using logical formulas; and a judgment result stored in the first storage device and a second storage device that stores the determination result stored in the first storage device. 1. A plant diagnostic device comprising: a diagnostic processing device that identifies a first cause of an abnormality using a cause-and-effect tree created by the operator, and outputs the first cause to a display device.