JP3056593B2 - Micro radiation leak detector - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a micro-radiation leak detecting device for detecting a micro-leakage of radiation to detect an abnormal occurrence of a nuclear reactor at an early stage.

[0002]

2. Description of the Related Art FIG. 4 is a block diagram showing a conventional micro-radiation leak detecting apparatus. In the figure, reference numeral 1 denotes a radiation source for radiation such as gamma rays or neutron rays, 2 denotes a radiation shielding material for shielding the radiation source 1, 2a is a minute gap of the radiation shielding substance 2, 3 is radiation leaked from the minute gap 2a or radiation existing in nature, 4 is installed outside the radiation shielding substance 2 and exists outside the radiation shielding substance 2 A radiation detector (for example, a plastic scintillator) that detects the radiation 3, 5 is a count integrator that integrates the radiation detected by the radiation detector 4 per unit time, and obtains a radiation amount per unit time, 6 is a predetermined alarm A setting device 7 for setting a value is an abnormality detection processing device for outputting an alarm when the radiation dose obtained by the count integrator exceeds an alarm value set in the setting device 6. A.

Next, the operation will be described. In a nuclear reactor, if a cooling water leakage accident occurs and the cooling capacity is reduced, there is a risk that accident propagation (for example, boiling of a coolant → fuel melting accident → adjacent fuel assembly) may occur. Therefore, when these abnormalities occur, it is necessary to detect the abnormalities at an early stage of the occurrence of the abnormalities and to take prompt measures, which is extremely important for the safety of the nuclear reactor.

Therefore, conventionally, a leak of radiation has been detected by using a minute radiation leak detection device shown in FIG. In detecting radiation leakage, first, the radiation detector 4 detects the radiation 3 existing outside the radiation shielding substance 2, and then the count integrator 5 converts the radiation 3 detected by the radiation detector 4 into unit time. And calculate the radiation dose per unit time.

[0005] Then, the abnormality detection processor 7 compares the alarm value set in advance by the setting unit 6 with the radiation dose per unit time determined as described above, and determines the radiation dose per unit time as the alarm value. If it exceeds, it is judged as abnormal and an alarm is output.

[0006]

Since the conventional micro-radiation leak detecting device is configured as described above, it is determined that an abnormality occurs only when the radiation dose per unit time exceeds a certain alarm value. Even if the radiation does not leak from the radiation shielding material, there is a small amount of radiation in the natural world, and the radiation dose is constantly fluctuating. There was a problem that it was not possible to distinguish whether the radiation was leaked or radiation originally existing in the natural world, and it was not possible to detect minute radiation leaked from the radiation shielding material.

SUMMARY OF THE INVENTION The present invention has been made to solve the above-described problems, and is capable of distinguishing between radiation leaked from a radiation shielding material and radiation originally existing in the natural world. It is an object of the present invention to obtain a minute radiation leak detection device capable of detecting minute radiation leaked from a radiation shielding material.

[0008]

According to the present invention, there is provided a micro-radiation leak detecting apparatus based on a radiation amount per unit time obtained by a count integrator at a normal time when radiation does not leak from a radiation shielding material. Predicting the radiation dose per unit time in normal time by creating a radiation autoregressive model, and detecting the minute leakage of radiation, the predicted value and the radiation dose calculated by the count integrator Then, an autocorrelation function of the error is obtained based on the error, and an alarm is output when the value of the autocorrelation function exceeds a predetermined alarm value.

[0009]

According to the present invention, there is provided a device for detecting minute radiation leaks.
An autocorrelation function of the error is obtained based on the error between the predicted value per unit time in normal time and the radiation dose obtained by the count integrator, and an alarm is issued when the value of the autocorrelation function exceeds a predetermined alarm value. By providing an abnormality detection processing device that outputs an alarm, it is possible to distinguish between radiation leaked from the radiation shielding material or radiation originally existing in the natural world, and when it is determined that the radiation has leaked from the radiation shielding material, an alarm is issued. Is output.

[0010]

【Example】

Embodiment 1 FIG. An embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing a minute radiation leak detecting apparatus according to an embodiment of the present invention. In the figure, the same reference numerals as those in the related art denote the same or corresponding parts, and a description thereof will be omitted. Reference numeral 8 denotes a band-pass filter, and 9 denotes an auto-regression model of the radiation 3 based on the radiation amount per unit time obtained by the count integrator 5 when the radiation 3 does not leak from the radiation shielding substance 2 in advance. In this way, a predicted value of the radiation dose per unit time in a normal state is obtained in advance, and when detecting a minute leak of the radiation 3, the predicted value is calculated based on the error between the predicted value and the radiation dose obtained by the count integrator 5. An abnormality detection processing device that obtains an autocorrelation function of an error and outputs an alarm when the value of the autocorrelation function exceeds a predetermined alarm value.

Next, the operation will be described. First, the radiation detector 4 detects the radiation 3 existing outside the radiation shielding material 2.
Is detected, the count integrator 5 integrates the radiation 3 detected by the radiation detector 4 per unit time, and calculates the radiation dose y _i (i = 1, 2,..., N) per unit time. At the same time, the radiation dose y _i is output to the abnormality detection processing device 9 through the band pass filter 8.

Next, the processing of the abnormality detection processing device 9 will be described. Since FIG. 2 is a block diagram showing the configuration of the abnormality detection processing device 9, it will be described with reference to FIG. First, the A / D converter 10 performs a process of converting the radiation dose _yi per unit time into a digital value. Next, the arithmetic unit 11
Are N time-series signals sampled at equal time intervals in a normal state where the radiation 3 does not leak from the radiation shielding substance 2, that is, the radiation dose y _i (i
= 1, 2,... N), an autoregressive model of the radiation 3 represented by the following equation (1) is created.

Y ″ _m = a ₁ y _m−1 + a ₂ y _m−2 +... + A _M y _mM ... (1) y ″ _m is a predicted value of y _m M + 1 ≦ m ≦ N A ₁ , a ₂ ,..., A _M in the equation (1) are autoregressive model coefficients determined so that the following equation (2) including the order M is minimized.

[0014]

(Equation 1)

[0015] Here, e _m is, N pieces of time-series signal y _m
When the formula (1) by the prediction value y of the calculated y _m "error between _m (hereinafter, referred to as error signal) is _{a. E m = y m -y"} m ··· (3) (m = M + 1 , M + 2,..., N)

The values of a ₁ , a ₂ ,..., A _M are obtained by the arithmetic unit 12 using the least square method, that is, the following arithmetic expression. The order M in the equation (1) is determined by an extremum search method. However, if the actual data is analyzed in advance and a guideline for the optimum order is provided, the calculation can proceed efficiently.

[0017]

(Equation 2)

Thus, a ₁ , a ₂ ,..., A
_When the value of _M is determined, the arithmetic unit 11 uses Expression (1) to calculate
The prediction value y ″ _m (m = M + 1, M + 2,..., N) is obtained.
This is performed as a preparatory stage before actually detecting the minute leakage of the radiation 3.

Next, when actually detecting a minute leak of the radiation 3, the subtractor 13 calculates the predicted value y ″ _m obtained by the calculator 11 and the radiation per unit time obtained by the count integrator 5. The error signal em with the quantity y _m is _expressed by the following equation (3).
Is determined using

Next, the correlator 14, on the basis of the error signal e _m, as described below, determining the autocorrelation function Φ _K ^ee (m).

[0021]

(Equation 3)

[0022] Here, FIG. 3 is the autocorrelation function Φ _K ^ee (m)
Is a waveform diagram showing a change in the value of the autocorrelation function. In a normal state where there is no radiation leakage, the autocorrelation function has a waveform as shown in FIG. It has the property of becoming a waveform like Therefore,
When there is no radiation leakage, the value of the autocorrelation function K is 1
When the radiation leak occurs, it rapidly attenuates and becomes lower at ~ 2 (it is theoretically guaranteed because the autoregressive model is determined so that equation (2) is minimized). It can be seen that the value of the autocorrelation function fluctuates greatly and does not attenuate even if the value of K is increased.

Therefore, utilizing this property, the setting device 15
At 18, an alarm value Ld1 and an alarm value Ld2 (error RMS
(See FIG. 3).
19 determines whether or not the following equations are satisfied. If either of them is established, it is determined that a radiation leak has occurred, and the alarm 17 or 20 outputs an alarm. ^{_{Φ 0 ee (m)> Ld2}} ··· (5) Φ K ee (m) ≦ -Ld1 or ^{_{Φ K ee (m) ≦ Ld1}} ··· (6) (K = 3,4, ···, j )

[0024]

As described above, according to the present invention,
Predicted radiation dose per unit time in normal time by creating a self-regression model of radiation based on radiation dose per unit time obtained by the count integrator at normal time when radiation does not leak from the radiation shielding material When detecting a minute leak of radiation, the autocorrelation function of the error is obtained based on the error between the predicted value and the radiation dose obtained by the count integrator, and the value of the autocorrelation function becomes Since it is configured to output an alarm when a predetermined alarm value is exceeded, it is possible to distinguish whether it is radiation leaked from the radiation shielding material or radiation originally existing in nature, and as a result, leaked from the radiation shielding material There are effects such as detection of minute radiation.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a minute radiation leak detecting apparatus according to an embodiment of the present invention.

FIG. 2 is a block diagram illustrating a configuration of an abnormality detection processing device.

FIG. 3 is a waveform chart showing a change in a value of an autocorrelation function Φ _K ^ee (m).

FIG. 4 is a block diagram showing a conventional minute radiation leak detection device.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Radiation source 2 Radiation shielding substance 3 Radiation 4 Radiation detector 5 Count integrator 9 Abnormality detection processor

Claims (1)

(57) [Claims] 1. A radiation detector installed outside a radiation shielding substance for shielding a radiation source, the radiation detector detecting radiation existing outside the radiation shielding substance, and the radiation detected by the radiation detector is transmitted every unit time. And a count integrator for calculating the radiation dose per unit time, and a radiation counter based on the radiation dose per unit time calculated by the count integrator when the radiation is not leaking from the radiation shielding material in advance. The predicted value of the radiation dose per unit time at normal time is obtained by creating an autoregressive model of the normal time, and when detecting a minute leak of radiation, the predicted value and the radiation dose obtained by the above-described counter integrator are calculated. Anomaly detection processing that calculates an autocorrelation function of the error based on the error of the error and outputs an alarm when the value of the autocorrelation function exceeds a predetermined alarm value. A micro radiation leak detection device comprising the device.