RU2548132C1 - Detection and identification of radioactive abnormalities in natural media and flows - Google Patents

Abstract

FIELD: instrumentation.

SUBSTANCE: one or several independent identical spectrometer detectors of gamma-radiation are used. These produce the common and cascade gamma-radiation by registration of time series of intensities including the primary gamma-radiation of radio nuclides for selected time of exposure. Factors of correlation of all elements of spectral composition between them are calculated in selected time interval. Correlation factors are standardised. Matrix of factors of pair correlations is constructed to detect and identify radioactive abnormalities. Matrix type allows identification of fluctuating radio nuclide while matrix change allows the definition of the start of fluctuation activity and its development in space or in time.

EFFECT: higher efficiency of detection of variations of natural medium radioactivity from tens of thousands of seconds to units of seconds, hence, higher accuracy of detection and identification.

1 cl

Description

The invention relates to the field of radioactive measurements.

Ecological monitoring of coastal waters and air of the driving atmosphere involves, among other things, control of the radioactivity field of natural environments. The purpose of monitoring is to obtain time series of field measurements of the concentrations of a more or less complete set of radioactive components of artificial and natural origin for the analysis of the spatiotemporal distribution of their variations (in this sense, the term “in stream” is used in the technique), trends and peak emissions and, on the basis of solving the background problem and the task of detecting and identifying, developing administrative, commander, research and other solutions. They are accepted in accordance with the developed criteria for the environmental situation, hazard criteria for the type of "environmental thermometer" (light bulb), the scale of which corresponds to the required solution.

First of all, this is a purely dosimetric problem of the radiation safety of the population, in this case the information obtained is a statistical time series of standard values of the exposure dose rate of gamma radiation at a height of 1 m above the surface. This is the control of emergency situations - monitoring of artificial radioactivity in the air and water environment by manifestations in the concentration field of its most representative representatives (Cs, I, etc.) This is the monitoring and study of the dynamics of air and water masses by manifestations of natural radioactivity in the field of concentration (K- 40, Rn, etc.)

As a rule, the characteristic scales of activity variations that interest the researcher are tens of meters or tens of seconds, i.e. variations of such scales should be detected and identified by the method.

A known standard method for recording gamma radiation of radionuclides "in situ" [1], adopted as a prototype of the claimed technical solution.

A radiation detector (which is a scintiblock based on a Nal (Tl) or CsI (Tl) single crystal, PMT, or PD, as well as power and signal transmission circuits placed in an airtight capsule) is placed in a medium, and spectrometric - in energy - line information communication is transmitted to the on-board recording device. To detect, identify and determine the statistical parameters of fluctuations in the component composition and activity of gamma emitters, the characteristic energy parameters of radiation are used, namely: counting time series in the energy range of the total absorption peak (SPP) of the gamma emitter, of interest to the researcher.

To do this, in natural conditions, periodically, measurements are interrupted, if possible, reference spectrometric radiation sources (SIRS) are introduced into the test medium, and the signal amplification path (radiation energy-channel number of the spectrometer) is calibrated using the IFR of reference radiation sources. Based on the calibration data, the operation of the spectrometric channel is monitored, the channel range of the spectrometer containing the SPP of the emitter under study is isolated, and parameters for the criteria for rejecting statistical information are obtained and, finally, the OSGI is removed and measurements continue. The procedure is lengthy, time-consuming and, to some extent, unsafe.

где τ - время экспозиции, α - коэффициент, соответствующий установленному статистическому критерию значимости, время экспозиции, необходимое для детектирования такого изменения активности (ΔΙ), составляет

и для 5%-ного уровня значимости равно 80 000 с, что значительно превышает анализируемые временные масштабы явления.In addition, if the background load (counting) of standard detectors in the SPP of natural gamma-ray emitters is ~ 1 pulse / s, and the natural fluctuations in their activity provide load additives ~ 0.01 pulse / s, then, based on standard significance criteria,

where τ is the exposure time, α is the coefficient corresponding to the established statistical significance criterion, the exposure time required to detect such a change in activity (ΔΙ) is

and for a 5% significance level it is equal to 80,000 s, which significantly exceeds the analyzed time scales of the phenomenon.

Overcoming these difficulties in ensuring the statistical reliability of measurements at short exposure times, for a standard measurement method, is possible by increasing the effective emitting volume of the medium supplying radiation to the detector by increasing its sensitive volume. However, in this case, the counting increases in proportion to the volume of the detector, and its cost due to technological reasons - in proportion to the cube of the volume.

A new approach to the analysis of the information content of the gamma field is required. Such a choice of recorded parameters and their analysis is necessary, which would make it possible to reduce the required threshold loadings during measurements and ensure the statistical reliability of the information received.

The basis of this approach for studying the gamma field are the features of the spectral composition of gamma radiation in a medium with distributed sources. First, the scattered radiation in the medium appears as a result of the Compton effect, the formation of electron-positron pairs and the photoelectric effect, from 4 to 8 quanta of softer energies are generated for each primary gamma quantum. Secondly, this emission spectrum in the medium appears in 10 ^-5 - 10 ^-7 s, which is significantly less than the exposure time, therefore, it has the property of appearance synchronization. And, finally, the decay of a number of radionuclides associated with the detection tasks is accompanied by cascade radiation with a characteristic time of 10 ^-7 - 10 ^-9 s, i.e. has the same synchronization property. It can be assumed that the analysis of the synchronism of counting fluctuations in separate ranges of the spectral composition of gamma radiation of the medium, as elements of the phase space (FP) of the gamma radiation field, when they are adequately normalized, will give a significant increase in the information content with respect to the problem posed.

The specificity of the field of radioactivity of natural environments is systematic. It lies in the fact that any impact on the natural environment causes adequate changes in many components of the radioactivity field, they are connected, systemic changes. This circumstance allows an environmental interpretation of the impact not according to the values of individual measured parameters, but also according to their interaction (systemic effect). Environmental factors - exposure parameters, are detected not by the measured parameters themselves (radionuclide concentrations), but by their systemic effect, even of the simplest type: in pairs. This concerns the direct impact, for example, the injection of a group of artificial radionuclides into the environment as a result of an atomic reactor accident through atmospheric deposition causes a systemic radioecological effect on human radiation safety through a number of food, inhalation and other chains. Similarly, the systemic effects of indirect factors, for example, a change in air mass with a different composition of natural radionuclides (marine or continental) as a result of general atmospheric circulation. Moreover, accurate accounting of all processes from the accident itself or processes in the upper atmosphere to the well-being of a person is practically impossible, and a system analysis of the “effect-effect” is quite feasible with obtaining real prognostic data of a statistical type.

The effectiveness of such an approach to solving problems associated with the fields of radioactivity of natural environments is known. For example, all methods of radioactive dating are based not on the analysis of the concentrations of individual radionuclides, but on a systematic analysis of the ratios of their concentrations, namely, on the simplest option - the analysis of pairwise concentration ratios, C-14 / C-12 or Pa-231 / Th-230. It is the concentration ratio that carries system information about the geological age. An impressive example of a systematic analysis of radioactivity in oceanology is the study of concentration ratios 0-18 / 0-16 to assess the parameters of regional ocean-atmosphere exchange [2]. A purposeful search for deviations from the standard for the ratio of the concentration of isotopes U-235 / U-238 made it possible to prove the existence of a nuclear reactor (natural nuclear reactor) in the territory of uranium mines in Gabon. While the individual concentrations of these isotopes, although they carry information about natural nuclear disasters, it is veiled by subsequent geological processes. In this regard, great hopes for confirming the adequacy of the NNR concept are assigned to systematic studies of the ratios of other pairs of radionuclides, the transformation of which can have a systemic effect during NNR processes, as well as other parameters of geological structures whose changes are associated with these processes [3]. The effectiveness of this approach was demonstrated by Georgievsky [4] for obtaining radioecological parameters in assessing the radiation situation after the Chernobyl accident by analyzing precisely the concentration ratios Cs-137 / Sr-90, Ru-106 / Ce-144, Zr-95 / Sr-90, Sr- 89 / Sr-90 et al.

In all these cases, system analysis was reduced to the simplest operation — the analysis of pairwise concentration ratios. Naturally, the question arises of a more general approach - the analysis of a multidimensional image of the phenomenon.

The system analysis procedure consists in conformally mapping a background multidimensional image of a field into an experimentally obtained image under the action of an event vector, in the general case. Depending on the method of formalizing a multidimensional image of the field, a system analysis tool is selected.

This situation is also inherent in the system of gamma-emitting isotopes of natural media, i.e. it also appears in the field of gamma radiation. The systemic effect in this case consists both in the fact that the “impact-tasks” cause certain, three related changes in the set of radionuclides injected into the environment, and in the fact that among them there is always a gamma-emitting component, which is determined by intensity and spectral composition, component of the gamma field.

Among other types of radiation, the gamma-field has, in this regard, several advantages: it contains complete information about the component composition of radioactivity, has a linear spectrum, an exceptionally high information transfer rate, and also optimal radiation transport parameters (the average mean free path in the medium exceeds typical design dimensions of the detectors, but less than the spatial scales of the processes under study). A rather simple way of detecting radiation, the high efficiency of registration with modern detecting systems compared to other types of radioactive radiation, and a wide range of nuclear-physical parameters of the emitters, which determines the high information content of the field for their identification, complements this series. Therefore, a systematic analysis of the statistical parameters of the gamma radiation field of the natural environment allows us to solve the indicated monitoring tasks, being, at the same time, the most promising way to automate the measurement process. And they mean by it the organization of a special configuration of the measuring complex, as well as the methodology for measuring and processing the obtained data, allowing them to be interpreted not only by the individual measured parameters, but by their combination, and in the general case throughout the multidimensional image of the field.

The basis of the invention is the task of creating a method for detecting and identifying radioactive anomalies in natural environments in a stream, the combination of the essential features of which achieves a new technical property - the ability to adequately select the elements of the gamma-field phase transition and system analysis of the counts in them. The specified new property provides the technical result of the invention - increasing the efficiency of statistically secured detection of variations in the radioactivity of the environment from tens of thousands of seconds to units of seconds, which increases the accuracy of detection and identification of radioactive anomalies.

The object of the invention is solved in that in the method of measuring and analyzing the statistical parameters of the gamma radiation field of natural media, in which one or more identical spectrometric gamma radiation detectors are used and the time series of the intensity of the primary gamma radiation of the radionuclide is recorded for the selected interval of time, new is that they record the time series of spectra of the general and cascade gamma radiation of radionuclides of the natural environment with a set time of position calculation is performed of the correlation coefficients of all elements of the spectral composition of each other at a selected time interval, producing a normalization correlation coefficients constitute a matrix of pairwise correlation coefficients for which detect and identify the radioactive anomalies.

The invention is illustrated as follows. They use detectors that make it possible to record gamma-ray spectra in a short exposure time relative to the scale of the analyzed phenomenon, which is an instant (in the sense of synchronism) image of the gamma-field by phase space elements. They include:

- the detecting system itself - a scintillation detector with a PM or PD, a preliminary signal amplifier and a power system;

- A pulse analyzer (AI), consisting of an amplifier, an amplitude-to-digital converter (ADC), and special computing equipment for constructing energy spectra of radiation, calibrating them, and transforming them to a convenient form;

- an AI communication interface with a more or less powerful computer;

- and, in fact, a computer, which allows one to obtain statistical conclusions and relationships for the time series of radiation spectra.

All positions starting from the second one in the above list can be replaced by a standard computer with the Matlab software package installed on it with the Simulink simulation program. In this case, the pulse signal from the preamplifier is input to the microphone input of the sound card, the internal ADC of which converts it to a digital code, and the corresponding discriminators, amplifier, and differential counters for 1024 channels are constructed in Simulink for standard spectrometry, the channel number of the 1024-channel spectrometer is recorded in which the pulse appeared and the current time of its appearance for pulse-by-wave recording of the signal. It is important to choose a sound card whose ADC dead time would be less than 0.5 mke, with an appropriate scale for digitizing the time stamp.

The software part of the complex consists of two software packages that differ in principle in purpose and, accordingly, in speed.

The first one contains a software package for receiving from the ADC - the signal amplitude code, and from the timer - the time stamp code of the arrival of a characteristic feature of the amplitude of the analog signal to the ADC (for example, upon reaching 0.7 amplitude) or the appearance of the code from the ADC and the write-through of the signal code to the operational memory. An important property of the signal recording device is speed, at the level of 0.5 μs. in this case, carries the head of this package. The average speed of analog signals from detectors is up to 10 pulses / s for real marine conditions, so the conditions of standard spectrometry can be easily provided with existing hardware. To ensure the operation of cascade radiation channels (the resolution time is not worse than 1 μs), it is necessary to introduce additional ADCs, one per channel, to register the code of signals from different channels by creating virtual microprocessors operating in parallel. The absence of a signal skipping in this case is ensured by a software pulse discriminator that rejects pulses with amplitudes of less than ~ 50 keV.

The second software package serves:

- to organize arrays of multidimensional time series of counts of the spectral composition of radiation of the natural environment;

- determination of the parameters of the spectral distribution of radiation of the medium and quality control of statistical information;

- detecting moments of significant changes in radiation intensity, highlighting a representative portion of the change, determining its magnitude and identifying the reasons for the variation of the gamma field;

- calculation of statistical parameters of the detection procedure and identification of the cause of the radioactive anomaly in the formalism of the probabilities of detection, signal skipping and false alarms.

The instrument for formalizing a multidimensional image of the gamma field is the correlation spectrometry apparatus. The complete information about the structure of synchronous fluctuations in the component activity is provided by the correlation matrix — the matrix of pair correlation coefficients (MCPC) compiled for all elements of the phase transition — a table Μ χ Μ of correlation coefficients.

и $$N_j^k$$

и коэффициент корреляцииThe time series (index k) of countings in the selected elements of the FP with numbers i and j is $$N_i^k$$

and $$N_j^k$$

and correlation coefficient

where K is the length of the implementation.

To build the confidence interval of the correlation coefficient, it is normalized using the Fisher transform [5]. Statistics

well approximated [6] by a normal distribution with parameters

In this case, the confidence interval for q has the form thZ, <Z <thZ2,

where Ν, is the normalized Laplace function for the l-% confidence interval. The values of Z ₁ and Z ₂ can be found in the tables, for example, [6].

The applicability of such a condition is ensured by the normality of N _i and N _j . In each case, it should be studied in a special way, in any case, this method gives reliable estimates for the score N> 30.

Moreover, the feasibility of this criterion is not necessary. Fisher analytically investigated the character of the distribution Ζ of q. An illustration of the results of these studies is presented in [6]. There are no differences in the distributions for q = 0 and q = 0.8 from the normal one. Beyond the scope of the research, зависимость remained dependent on the nature of the distribution of значений values, in the elements of the phase transition.

и

Для моделирования значений рядов применялся метод Неймана, что позволяет ожидать некоррелированность результатов выборки для 106 событий. Результаты моделирования подтвердили предположение о независимости распределения Ζ от величины загрузок Ν, вплоть до единиц импульсов в элементах ФП. Для модельных рядов

при длине реализаций 103 отклонения от нормальности не зависят от Ν_i и Ν_j и для выборок ~10 000 не превышают 0,5%.The numerical simulation of the probability density of the distribution Ζ for various Ν having a Poisson distribution was carried out by the Monte Carlo method by direct modeling of two Poisson series consisting of sections with

and

To simulate the values of the series, the Neumann method was used, which allows us to expect uncorrelated sampling results for 106 events. The simulation results confirmed the assumption that the distribution Ζ is independent of the load Ν, up to units of momenta in the elements of the phase transition. For lineups

for a realization length of 103, deviations from normality are independent of Ν _i and Ν _j and for samples of ~ 10,000 do not exceed 0.5%.

Thus, the minimum spatio-temporal scales of the variability of the structure of the gamma field of the natural environment, which can be statistically reliably identified by the proposed method, are determined by threshold loads in the elements of the gamma field phase transition that do not exceed units of pulses during the exposure time.

To make decisions appropriate to the ecological situation, it is necessary to convert the generalized multidimensional ecological image of the environment into a simple one-dimensional generalized parameter or a simple flat picture, the elements of which reflect a particular state of the environment (ecological status) and its temporal trajectory. The one-dimensional image actually represents a certain scale of the state of the environment, selected depending on the requirements, or "environmental thermometer." Its division price is set by some reference states of the environment obtained by calculation or experimentally, including the background state. The mathematical tool for such a mapping is an apparatus for analyzing multidimensional time series with all its components: the principal component analysis, cluster analysis, factor analysis, nonlinear scaling, simulation statistical experiments based on various randomization methods (for example, budstrap operations).

It is important to note that it is a multidimensional image (in our case, MKPK) that contains the maximum information about the ecological state of the environment, and any display of it in one-dimensional or flat images emasculates this information content. A criterion for the adequacy of information content and, accordingly, the effectiveness of the display procedure (selection of the ecological thermometer scale) is the assessment of its sensitivity and the possibility of solving the inverse problem.

The method is carried out, for example, as follows.

Three identical gamma-ray spectrometric radiation detectors, which are independent of each other, are placed in the medium in hardware. Spectra of total and cascade gamma radiation are obtained, including by recording the time series of the intensity of the primary gamma radiation of radionuclides for a selected time interval. Carry out, for example, using relations (1), (2) calculate the correlation coefficients of all elements of the spectral composition with each other on a selected time interval. For example, correlation coefficients are normalized using relation (3). A matrix of pair correlation coefficients is compiled, according to which radioactive anomalies are detected and identified - the type of the matrix identifies the fluctuating radionuclide in the selected time interval, and the beginning of the fluctuation activity and its development in space or time is determined by the change in the matrix in time.

For the optimal use of statistics, nonlinear capping of spectrometric paths is introduced. The recorded energy spectrum has an exponential form, therefore, soft energy channels have an excess of counting, compared with the above criterion of statistical security, at the same time, in hard energy channels (more than 2 MeV), the count is no more than 0.1-0.01 imp / s . To align channel-by-channel statistical security, a nonlinear energy calibration of spectrometric paths is introduced in software, which ensures uniform channel-by-channel path loading. This allows, on the one hand, to optimally use the available downloads, and on the other hand, it increases the number of spectrometer channels in the soft part of the spectrum, where the identification capabilities of the method by SPP are significantly reduced.

To carry out the calibration of spectrometric paths during measurement in the medium, continuous calibration can be carried out using the SPP of natural emitters (K-40, Bi-214, Pb-214). This eliminates the cost of an OSGI calibration procedure without interrupting the measurement process.

To correct the swimming of the gain of the spectrometric path, a continuous floating calibration of the spectrometric path and, accordingly, the correction of the channel boundaries of the spectrometer can be programmed.

Programmatically between identical and independent channels, cross-counting of counting fluctuations can be carried out. By the synchronization of the path response, the time series for the channels are adjusted or rejected based on the selected significance criteria.

The measuring complex can additionally include analogue measuring channels (temperature, pressure, sound velocity, electrical conductivity, meteorological parameters, etc.), which, unlike quantum radioactive ones, can in some cases increase the reliability of the detection, identification, and determination of statistical parameters, in others, they serve as a criterion for the reliability of identification of fluctuations by radioactive parameters.

To control the sample in order to process statistical information but adjusted time series in connection with a change in the calibration parameters of the paths, a pulse-by-pulse recording of spectrometric information can be introduced. This allows us to introduce more subtle methods of processing statistical information at the stage of analysis of a multidimensional image of the radiation field.

Information sources:

1. B.V. Grinev, V.D. Ryzhikov, V.P. Seminozhenko. Scintillation detectors and radiation monitoring systems based on them. Kiev, Naukova Dumka, 2007, S. 451.

2. V.Η. Eremeev, B.N. It's ridiculous. Oxygen-18 as an indicator of regional exchange processes in the ocean-atmosphere system. Physical Aspects of Atmospheric Pollution - Vilnius - 1976. T. 2. - P. 225-227.

3. I.F. Lukashin. The concept of nuclear power. In the book. Environmental control systems. Sevastopol, 2007 .-- S. 311-313.

4. V. B. St. George's. Environmental and dose models for radiation accidents. Kiev. Naukova Dumka. 1994 .-- 235 p.

5. Fisher R.A., Biometrika, 10, (1915), 507.

6. Hudson D. Statistics for physicists. M., ed. World, 1970, p. 296.

Claims (1)

 A method for detecting and identifying radioactive anomalies in natural media in a stream, which consists in using one or more identical spectrometric gamma radiation detectors that are independent of each other and registering a time series of the intensity of the primary gamma radiation of the radionuclide for a selected time interval, characterized in that register the time series of spectra of the total and cascade gamma radiation of radionuclides of the natural environment with a specified exposure time, calculate the coefficients correlations of all elements of the spectral composition with each other at a selected time interval, normalize the correlation coefficients and compose a matrix of pair correlation coefficients, which detect and identify radioactive anomalies.