JPH0797436B2 - Power plant transient data display - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Object of the Invention] (Field of Industrial Application) The present invention relates to a process signal display device when an abnormality occurs in a large-scale power plant.

(Prior Art) There are many abnormality diagnosing devices for large-scale power plants such as nuclear power plants, but among these, diagnosis at the time of transient change due to equipment failure (quick avoidance operation and cause investigation are required). Is important and many suggestions have been made. Transient changes in a large-scale system have complex and intricate behavior of many process signals (pressure, flow rate, etc.), so the data observed by conventional instrumentation systems can quickly and accurately identify the cause of the abnormality. Is difficult.

As a method of diagnosing such an instrumentation system, a cause-causal tree (CCT) between process signals is prepared, and the cause of abnormality is automatically estimated by matching with the observation pattern, or when the plant is operating normally. As a typical diagnostic method, a model is prepared and the cause of abnormality is identified from the disagreement between the prediction from this model and the observed value. However, these diagnostic methods are based on the assumption that the plant CCT and model are sufficiently accurate. In practice, the accuracy of these models depends on, among other things, the type of anomalous change, so there can be false positives. Therefore, relying only on the CCT and the model, trusting the diagnostic results from them, and considering the subsequent actions, is considered dangerous if the diagnostic results are incorrect. In such cases, it is necessary to provide operators with accurate information on the plant situation based on observations that are as unpretentious as possible.

The method usually used to provide such observation data is trend display using a trend recorder or computer, or a pie / bar graph display. However, many signals, such as abnormal transient changes, are complicated and complex at the same time. In some cases, it is not very suitable for displaying when it changes due to entanglement.
Simultaneous computer-aided plotting of multiple signal trends is sometimes useful if the relevant signals are collected and displayed, but the type of anomaly may make it inappropriate for plotting a fixed group of signals. .

(Problems to be Solved by the Invention) The present invention has been made in view of the above points, and an object of the present invention is to provide an operator with an optimal behavior in the event of an abnormal transient change in which many process signals are intricately entangled. Is to display.
In particular, it is an object of the present invention to provide a power plant transient data display device that considers the cause-effect causal relationship in accordance with the type of anomaly that occurs and rearranges and displays the signals in a form that follows the causal relationship.

[Structure of Invention]

(Means for Solving the Problem) In order to achieve the above object, the power plant transient data display device of the present invention is an A / D converter for continuously A / D converting the process signal of the power plant, An abnormal transient change detector that detects abnormal transient changes in the process signal from changes in the digital value of the A / D converter, a data storage device that saves time-series data for several minutes before and after the abnormal detection, and an abnormal transient change Causal relationship network between process signals in which the time series data of the process signals are registered in advance, and three scales for all detected signals of abnormal transient changes: those with early detection, those with large transient changes and important ones. A calculation that creates a causal tree in which the output is rearranged according to the importance of the signal from this selected signal by calculating the weight with A storage device and a display device for sequentially displaying the output of the calculation / storage device.

(Operation) Therefore, according to the power plant transient data display device of the present invention, the behavior of a complicated process signal at the time of an abnormal transient event occurring in a large-scale plant can be detected by using the original trial / causal relationship network and physically. It can be output along the propagation path.

Next, an explanation will be given of the process data display device at the time of abnormal transient change of the present invention with reference to FIG. The present invention uses a cause-effect causal network between process signals as shown in FIG.

FIG. 2 is a directed diagram showing the relationship between cause and effect between process signals. For example, when the dome pressure rises, the main steam flow rate increases (→ arrow), and when the main steam flow rate increases, the dome pressure decreases (← arrow).

Next, if an abnormal transient change is detected by some signal,
The signal of interest is weighted on three scales: one that detected earlier (1 / t ₁ ), one that has a large transient change (α _i ) and one that is important (β _i ). Is calculated for all the signals (i) in which a change is detected, and is selected as the maximum. Where t _i is Fi
Relative detection time from the detection time of rst Hit, α _i is the magnitude of change, β _i is a value preliminarily held as the importance of the signal,
a, b, c, △ are appropriate weights. Of course, the most noticeable signal can be selected by changing the combination of the other scales.

In this way, when a signal of interest is selected, a causal tree is created by developing directed graphs on the upstream and downstream sides of the arrow centered on that signal. Fig. 3 (A) is obtained by expanding the directed diagram on the upstream side of the arrow centering on the signal of interest (here, the main steam flow rate) from Fig. 2, and similarly, the directed diagram on the downstream side of the arrow. Is developed to obtain FIG. 3 (B). This FIG. 3 (A)
In both figures (1) and (B), numbers are assigned in the order of left-to-right, depth order (horizontal search) (from Figure 3), and up to the number (MAXN) that can be displayed on one screen. Leave only the trees and cut the others. The two kinds of trees created in this way are
FIG. 4 (A) is obtained by ordering depths from left to right (vertical search) up to and displaying them all together on the screen in this order. Of course, if the resolution of the CRT allows it, it is possible to display it in a tree shape as shown in FIG. 4 (B).

In the above causal tree development, for example, there may be an infinite loop such as main steam flow rate ← dome pressure ← main steam flow rate ← ..., but this can be easily avoided by loop check.

Further, if there is an abnormality detected in addition to the signals displayed above, the equation (1) is applied to the remaining abnormal signals excluding the above-mentioned display signals to select the signal of interest and display in the same procedure. It is also possible to do so.

(Example) The Example of this invention is described about drawing.

FIG. 1 is a block diagram of an embodiment of the present invention. In the figure, a power plant transient data display device 1 of the present invention is an A / D converter 2 for converting an analog signal into a digital signal.
An abnormal transient change detecting device 3 for detecting an abnormal transient change in a process signal from a change in a digital signal; a data storage unit 4 for storing data in time series when an abnormal condition is detected; The operation / storage unit 5 arranges the data in the output order according to the degree of importance of the signal, and the display device 6 for displaying the output of the operation / storage unit 5.

Next, the plant transient data display procedure of this embodiment will be described.

First, a plurality of process signals are continuously taken in by the A / D converter 2 from the plant.

Next, the abnormal transient change detection device 3 detects an abnormality by the following algorithm. First, x (t) (t = Δt, 2Δ
t, ...) is the time series data of a specific process signal.
Next, the moving average is calculated by the following equation (2).

Here, T is a time constant, Δt is a sampling period, and y
(T) is a moving average value. Further, the abnormality is detected by | x (t) −y (t) |> nσ ... (3). Here, σ is the standard deviation of the signal in the normal state, and the deviation from the normal state can be detected sensitively by detecting the abnormality with a threshold value n times that.

When an abnormality is detected in this manner, K data before and after the abnormality is stored in the data storage device 4.

Further, in the calculation / storage device 5, the output sequence is automatically set according to the causal network between the process signals registered in advance and the time series data of the stored process signals at the time of the abnormal transient change and the importance of the signals. After rearranging, they are displayed by the display device 6 in the order.

〔The invention's effect〕

As described above, the behavior of the complex process signal at the time of abnormal transient event occurring in the large-scale plant according to the present invention,
It is possible to output along a physical propagation path using a causal / causal network. Furthermore, by selecting the signals of interest and creating a causal tree, the output order is rearranged according to the importance of the signals. The effect is that it is possible to respond accurately.

[Brief description of drawings]

FIG. 1 is a block diagram of one embodiment of the present invention, FIG. 2 is a directed diagram of causal relationship propagation of process signals according to the present invention, and FIGS. 3A and 3B are causes according to the present invention. FIG. 4 (A) and FIG. 4 (B) are development trees in a direction and a result direction, and are diagrams showing a method of displaying time series data along a propagation path according to the present invention. 1 ... Power plant transient data display device 2 ... A / D converter 3 ... Abnormal transient change detection device 4 ... Data storage device 5 ... Calculation / storage device 6 ... Display device

Claims (1)

[Claims] 1. A process signal of a power plant is continuously A /
An A / D converter that performs D conversion, an abnormal transient change detection device that detects abnormal transient changes in process signals from changes in the digital value of this A / D converter, and time series data for several minutes before and after the abnormal detection A data storage device to save, a causal relationship network between process signals in which time series data of process signals at the time of abnormal transient changes are registered in advance, and all detected signals of abnormal transient changes, those that were detected earlier, transient changes Is calculated by weighting on three scales, the large one and the important one, and the maximum one is selected as the signal of interest, and the output order is rearranged from this selected signal according to the importance of the signal. A power plant transient data display device comprising a calculation / storage device for producing fruit trees and a display device for sequentially displaying outputs of the calculation / storage device.