Radiometric limit switch and method for level monitoring using the radiometric limit switch

Abstract

Radiometric limit switch, with - a spotlight (5), ◯ sends radiometric radiation along a radiation path running at the height of the limit level to be monitored, or which is in a free state when the limit level has been undershot, and or which is in a state covered by a medium when the limit level is exceeded, - a detector (7) inserted at the end of the beam path, which is designed to measure discrete radiation intensities (I) arriving at it, which depend on the state of the radiation path, wherein a statistical distribution of the radiation intensities (I) in the free and in the covered state is given by two separate Poisson distributions, - a measuring instrument electronics (15) connected to the detector (7), which ◯ an interval (T) starts in which a successive distribution (V(T)) of the measured radiation intensities (I) is recorded, o the distribution (V(T)) is continuously or at predetermined time intervals examined to see if it contains two separate Poisson distributions, o the interval (T) ends after two separate Poisson distributions have been identified in the distribution (V(T)) recorded up to that point, o based on the positions of the two Poisson distributions identified in the distribution (V(T)) within the distribution (V(T)) an upper threshold (S max ) and / or a lower threshold (S min ) for the radiation intensity (I), and o those in case of exceeding or falling below the threshold value (S min , S max ) detects a change of state to the free or covered state by measuring the radiation intensity (I) following the interval (T).

Description

[0001] The invention relates to a radiometric limit switch and a method for monitoring the level, comprising a radiometric limit switch with a emitter that emits radiometric radiation along a radiation path running at the level of the level to be monitored, which is in a free state when the level to be monitored is undershot, and which is in a state covered by a medium when the level to be monitored is exceeded, and a detector inserted at the end of the beam path that measures discrete radiation intensities arriving thereon, which depend on the state of the radiation path, and in which an exceedance or undershooting of the level to be monitored is monitored on the basis of the measured radiation intensities.

[0002] Radiometric limit switches are typically used when conventional limit switches cannot be used due to particularly harsh conditions at the measuring point. Very often, for example, extremely high temperatures and pressures prevail at the measuring point, or there are chemically and / or mechanically very aggressive environmental influences present, which make the use of other measuring methods impossible.

[0003] Radiometric level switches are used, for example, to monitor whether a predetermined fill level of a substance in a container is exceeded or fallen below. They serve, for instance, as overfill protection or as dry-run protection.

[0004] For this purpose, they feature a radioactive source mounted externally on the container, which, during operation, emits radioactive radiation along a radiation path running through the container at the level of the limit to be monitored. On the opposite side of the container, a detector is mounted externally, inserted at the end of the radiation path, which quantitatively measures the radiation intensity exiting the container. The exiting radiation intensity depends on the geometric arrangement and the absorption. The latter depends on the fill level of the material in the container and its density.

[0005] Accordingly, the radiation intensity exhibits a minimum, dependent on the density of the contents, when the fill level is above the radiation path. Conversely, the radiation intensity exhibits a maximum when the fill level is below the radiation path.

[0006] The minimum and maximum radiation intensity are now regularly determined in a balancing procedure to be carried out by the user during the commissioning of the radioactive limit switch, in which the two above-mentioned filling states are set freely and covered in the container, and the corresponding minimum and maximum radiation intensities are measured by the limit switch.

[0007] Based on the minimum and maximum radiation intensities, a threshold value for the radiation intensity, usually referred to as the switching point, is defined. Exceeding this threshold corresponds to a change to the unobstructed state, and falling below it corresponds to a change to the obscured state. During subsequent operation of the limit switch, a comparison of the measured radiation intensity with the threshold value determines whether the level to be monitored has been exceeded or fallen below. A corresponding radiometric limit switch is described, for example, in German patent DE 11 72 864 B1. To achieve the most stable switching behavior possible, a switching hysteresis is often incorporated. For this purpose, a lower and an upper threshold value are defined based on the minimum and maximum radiation intensities.In operation, a change to the unobstructed state is only indicated when the measured radiation intensity exceeds the upper threshold, and a change to the obscured state is only indicated when the measured radiation intensity falls below the lower threshold. In the range between the two thresholds, the limit switch continues to output the last determined state. Calibration procedures now regularly require the user's assistance, who must establish the necessary conditions in their system for calibration and indicate the presence of each state to the limit switch so that it can measure the corresponding maximum and minimum radiation intensities and store them, assigning them to the respective state.The setting of the two filling states, free and covered, is usually associated with an impairment, interruption and / or delay of the manufacturing and / or processing procedure or an ongoing process taking place at the measuring point.

[0008] It is an object of the invention to provide a method for radiometric level monitoring in which the limit switch automatically initiates the monitoring operation, in particular without prior execution of a calibration procedure in which at least one of the states must be established and the associated radiation intensity measured.

[0009] The invention comprises a radiometric limit switch according to the features of independent claim 1 and a method for limit level monitoring with the radiometric limit switch according to the features of claim 3. Advantageous embodiments thereof are described in dependent claims 2 and 4 to 9.

[0010] The invention and further advantages will now be explained in more detail with reference to the figures in the drawing, in which two exemplary embodiments are shown; identical parts are provided with the same reference numerals in the figures. Fig. Figure 1 shows: a measuring arrangement with a radiometric limit switch; Fig. 2 shows: the detector of Fig. 1 and a connected measuring electronics unit; Fig. Figure 3 shows: a Poisson distribution of the pulse rates resulting in the free state and a Poisson distribution of the measured pulse rates resulting in the covered state; and Fig. Figure 4 shows: the recording of a distribution of the measured radiation intensities in a histogram.

[0011] Fig. Figure 1 shows a measuring arrangement with a radiometric limit switch attached to a container 3 filled with a substance 1 for monitoring whether a predetermined limit level is exceeded or fallen below. The limit level here is a predetermined fill level L of the substance 1 in the container 3. The container 3 is, for example, a tank, a container, a pipe, a conveyor belt, or any other container shape.

[0012] The radiometric level switch comprises a radioactive source 5 that emits radioactive radiation along a radiation path running at the level of the level to be monitored – shown here as a dashed line. In the illustrated embodiment, the source 5 is mounted externally at the level of the predetermined fill level L in a radiation protection container on the container 1. The source 5 is, for example, a gamma radiation source, such as a cobalt-67. 60 or Cs 137Preparation. However, the invention can also be used in conjunction with other types of radiation sources, such as neutron emitters.

[0013] The radiation protection container has a recess through which the radiation emitted by the source 5 penetrates through the container 1 along the radiation path at the level of the limit to be monitored.

[0014] Depending on the fill level in container 3, either a state referred to below as the "free state" exists, in which the fill level is below the limit level to be monitored, or a state referred to below as the "covered state" exists, in which the fill level is above the limit level to be monitored. In the free state, the radiation path in container 3 is unobstructed. In the covered state, the radiation path in container 3 is covered by a medium, in this case, the contents 1.

[0015] On the side of the container 3 opposite the emitter 5, a detector 7 is provided on the outside, inserted at the end into the radiation path, which successively detects discrete radiation intensities I(t) arriving on it, depending on the state of the radiation path. i ) measures, each corresponding to a number of radiometric radiation quanta arriving on it per unit of time Δt.

[0016] A suitable detector 7 is, for example, one that is in Fig. Figure 2 shows a detailed scintillation detector with a scintillator 9, e.g., a scintillation rod, and an attached light receiver 11. The scintillator 9 is made of a special plastic, such as polystyrene (PS) or polyvinyltoluene (PVT), which is optically very pure. Under the influence of gamma radiation, flashes of light are emitted by the scintillation material. These are detected by the light receiver 11 and converted into electrical pulses k. The light receiver 11 is, for example, a photomultiplier. Alternatively, avalanche photodiodes or semiconductor detectors, e.g., cadmium-zinc-telurd semiconductor detectors, can also be used. A measuring electronics unit 15 is connected to the photomultiplier 11 via a pulse line 13. Based on the incoming pulses k, the electronics unit 15 uses a counter 17 and an internal clock 19 to determine the pulse rate N(t). i ) determined. The impulse rate N(t i) is equal to the number n of pulses k detected per unit time Δt, and thus a direct measure of the radiation intensity I(t) arriving at detector 7 i ). Impulse rate N(ti) and radiation intensity I(t) i These are equivalent measurands that can be directly converted into one another. Alternatively, other types of detectors, such as Geiger-Müller counter tubes or ionization chambers, can be used, in which the incoming radiation is also converted into flashes of light, which are then converted by a light receiver into an electrical signal representing the incoming radiation intensity I.

[0017] The radiation intensity I arriving at detector 7 depends on the radiation power of the emitter 5 and the absorption along the radiation path from emitter 5 to detector 7. While the absorption in the container walls is constant in the application, the absorption along the path d traveled in container 3 depends on the density ρ and the attenuation coefficient µ of the medium in the radiation path within container 3.

[0018] In the free state, the radiation intensity I exhibits a maximum I max that the initial radiation power I0 is given by the radiation power of the emitter 5 and the absorption in the container walls, and an exponential dependence on the density ρ0 of the gas in the radiation path, usually air, the length of the path d traveled through the gas of the radiation and the attenuation coefficient µ0 of the gas. Imax=I0e−μ0ρ0d

[0019] Accordingly, the radiation intensity I exhibits a minimum in the cloudy state. min that the output radiation conductance I0 and an exponential dependence on the density ρ L of the filling material 1 located in the radiation path, the length of the path d traveled by the gas of the radiation and the attenuation coefficient µ L of the filling material 1 is given. Imin=I0e−μLρLd

[0020] Due to the quantum nature of radiometric radiation, the measured radiation intensities I(t) i ) discrete measured values ​​that occur in the two states in a statistical distribution that can be very well described by the static distribution of the measured pulse rates N(t i ), i.e., the number n of impulses k arriving per unit of time Δt, can be illustrated.

[0021] If the radiation path is in one of the two states 'free' or 'covered', then the probability p(N) of the occurrence of a certain pulse rate N is given by a Poisson distribution according to: p(N)=1N!(N¯)Ne−N¯ given, where N in each state is equal to the mean of the pulse rates N occurring in that state.

[0022] Fig. Figure 3 shows the distribution p resulting in the free state. f (N) and the resulting distribution p in the covered state b (N). The two distributions p f (N) and p b (N) each exhibit a pronounced maximum at the mean value N f , N b the respective distribution p f (N) and p b (N) on. The mean N f the distribution p resulting in the free state f (N) corresponds to the mean radiation intensity I expected in the free state. max The mean N bthe distribution p resulting in the covered state b (N) corresponds to the significantly lower average radiation intensity I expected under cloud cover. min .

[0023] The two distributions p f (N) and p b (N) are two clearly separated Poisson distributions whose width is each given by the variance σ. f , σ b is given by the respective Poisson distribution, which is equal to the square root of the mean N f , N b the respective distribution p f (N) and p b (N) is. Therefore: σf=Nf¯ and σb=Nb¯

[0024] The distance between the two Poisson distributions is determined by the differing absorption in the two states. Level monitoring with the inventive method, just as with conventional methods, is only feasible if the difference in absorption between the two states is large enough to produce clearly distinguishable intensity measurements. This necessarily means that the two Poisson distributions must be separated and thus distinguishable. The minimum distance between the two Poisson distributions required for them to be considered separate can be determined, for example, using integer half-value layers H, by the following condition: Nf¯Nb¯≥2H A minimum spacing of at least 5 half-value layers, i.e., H = 5, is typically required. The user can usually ensure compliance with this condition by estimating it based on their knowledge of the application.

[0025] The statistical properties and relationships mentioned above, based on Poisson distributions, apply due to the equivalence of radiation intensities I(t). i ) and pulse rates N(t i ) naturally in the same way for the radiation intensities I(t) i According to the invention, these statistical properties and relationships are exploited to enable automatic commencement of monitoring operation by the limit switch, which in particular does without prior execution of a calibration procedure in which at least one of the states must be established and the associated radiation intensity measured.

[0026] For this purpose, detector 7 is used after the installation of the limit switch in an interval T = [t0..., t m ], in which the two states free and covered are assumed at least once, successive radiation intensities I(t0), ... I(t m ) measured, and a distribution V(T) of the radiation intensities I(t) measured in the interval T i ) recorded. It is completely irrelevant which of the two states the radiation path is in at the beginning of the interval T. In particular, it is not necessary to pre-set the radiation path to a specific state or to know its current state.

[0027] Subsequently, two separate Poisson distributions are identified within the recorded distribution V(T), and an upper threshold S is determined based on the positions of the Poisson distributions identified in the distribution V(T). maxfor the radiation intensity I, the exceedance of which is determined by a radiation intensity I measured after the interval T, resulting in a change of state to the free state, and / or a lower threshold S min for the radiation intensity I, below which a change of state to the overcast state is detected by a radiation intensity I measured after the interval T.

[0028] The procedure can, of course, be carried out analogously using the corresponding pulse rates N. Pulse rates N and radiation intensities I are used synonymously here due to their equivalence.

[0029] According to a first embodiment of the invention, the interval T is a period prior to the commencement of monitoring operation of the limit switch, during which both states occur at least once. In this embodiment, the limit switch automatically begins to successively measure radiation intensities I(t) immediately after being switched on at time t0. i ) to measure. The distribution V(T) of these measured radiation intensities I(t) i ) is recorded, for example, in the form of a histogram, in which the frequencies # with which the individual discrete radiation intensity measurements occur are recorded. Fig. Figure 4 shows the creation of such a histogram, in which the frequencies # are plotted as a function of the discrete measured radiation intensity values ​​I. This procedure step is carried out at least until two separate Poisson distributions can be identified in the recorded distribution V(T). This is done, for example, by means of an electronic unit 23, e.g., a microprocessor, integrated into the measuring instrument electronics 15 and connected to a memory 21.

[0030] Meanwhile, the recorded distribution V is continuously examined, or, as in the example shown here, at predetermined time intervals T1, T2, T3 ... by appropriate software to determine whether the distribution V(t) recorded up to that point already exhibits two separate Poisson distributions.

[0031] This can be done, for example, by determining the maxima M of the frequencies # of the distribution V(t) recorded up to that point. Preferably, a minimum frequency # is also considered. min specified values ​​that must exceed the identified extreme values ​​of the frequencies # in order to be considered maxima M of the distribution.

[0032] In the Fig. In the example shown in section 4, the radiation path was initially in the free state. From the start of the recording at t0 until time T1, the recording yielded the distribution V(T1). It exhibits a single minimum frequency # min exceeding maximum M1.

[0033] As long as the recorded distribution V(t) has only one maximum, here M1, two separate Poisson distributions cannot be identified, and the procedure step continues.

[0034] The next in Fig. The snapshot shown in section 4 depicts the distribution V(T2) recorded up to time T2. This distribution V(T2) exhibits two different minimum frequencies. min exceeding maxima M2 and M3.

[0035] As soon as two different ones meet the minimum frequency # min Exceeding maxima, here M2, M3, are present, is determined based on the associated radiation intensity measurements I max2 and I max3 It was checked whether these could be assigned to a single or two different Poisson distributions. This is done, for example, by assuming that the two radiation intensity measurements I max2 and I max3 are approximately equal to the mean values ​​of the corresponding Poisson distribution. Accordingly, one of the two radiation intensity measurements I is used. max2 and I max3 , preferably based on the larger of the two, here I max3, a measurement interval Δ(M3) is estimated within which the intensity measurements corresponding to this Poisson distribution are expected. Since the variance σ f , σ b If the value of a Poisson distribution is equal to the square root of its mean, the measurement interval Δ(M3) can, for example, be replaced by a measurement value I that is larger than the radiation intensity measurement. max3 a centered interval can be estimated, the width of which is determined on both sides of the radiation intensity measurement I. max3 greater than or equal to twice the square root of this radiation intensity measurement I max3 is.

[0036] If the intensity measurement value I max2 of the other maximum M2 of the recorded distribution - as represented in the distribution V(T2) - within this measurement interval Δ(M3) = [I max3 - 2 (I max3 ) 1 / 2 , I max3 + 2 (I max3 ) 1 / 2If the radiation path is ], then it can be assumed that it belongs to the same Poisson distribution. The limit switch automatically recognizes that the process step must continue, since the radiation path is either still in its initial state, or the other state has not been present long enough to identify its Poisson distribution.

[0037] The next in Fig. The snapshot shown in section 4 depicts the distribution V(T3) recorded up to time T3. This distribution V(T3) exhibits three different minimum frequencies. min The maximum values ​​M4, M5, and M6 exceed the limit. Here again, it is checked whether the maximum values ​​M4, M5, and M6 can be assigned to two separate Poisson distributions, preferably using the highest intensity measurements I. max6 and I max5 is started. The intensity measurement I max5 lies within the measurement interval Δ(M6) = [I max6 - 2 (I max6 ) 1 / 2, I max6 + 2 (I max6 ) 1 / 2 ] to obtain the highest intensity measurement I max6 . Accordingly, these two maxima M5, M6 are assigned to one and the same Poisson distribution.

[0038] The intensity measurement I max4 lies outside the measurement intervals Δ(M6) = [I max6 - 2 (I max6 ) 1 / 2, I max5 + 2 (I max6 ) 1 / 2 ] and Δ(M5) = [I max5 - 2 (I max5 ) 1 / 2 , I max5 + 2 (I max5 ) 1 / 2 ] to determine the two highest intensity measurements I max5 and I max6 Accordingly, the limit switch now automatically identifies the presence of two separate Poisson distributions. From this, the limit switch recognizes that both states of the radiation path have occurred at least once. The interval T in which the distribution V(T) of the measured radiation intensities I(t) i ) being recorded can therefore be terminated.

[0039] The position of each of the two identified Poisson distributions is now determined. For this purpose, the mean values ​​of the two identified Poisson distributions, around which the Poisson distributions are centered, are preferably determined, and the position is set equal to the respective mean value. Subsequently, based on the mean value of the Poisson distribution centered around a lower mean radiation intensity in the distribution V(T), an expected mean radiation intensity I under cloud cover is calculated. b determined, and based on the mean of the Poisson distribution centered around a higher mean radiation intensity in the distribution (V(T)), an expected radiation intensity I in the free state is calculated. f certainly.

[0040] Based on the two mean values ​​and their assignment to the two states of the radiation path, an upper threshold S is now determined. maxfor the measured radiation intensities I, above which a transition to the free state is detected, and a lower threshold S min for the measured radiation intensities I, below which a change to the cloudy state is observed.

[0041] The two threshold values ​​S min , S max are each placed on a point between the two mean radiation intensities I f , I b The lying value is set.

[0042] The determination of the two threshold values ​​S min , S max This can be done in the same way as with conventional limit switches, where the average radiation intensities expected in the free and covered states I f , I b have been determined in a preliminary comparison procedure.

[0043] Since the means of the two Poisson distributions determine not only their position, but also their variances σ(I) f ) and σ(I b ), which are each equal to the square root of the corresponding mean, and whose width is also known, the threshold values ​​S can be determined. min , S max now optimally configured for the desired application.

[0044] For example, to achieve the largest possible hysteresis, the distance between the two threshold values ​​S can be as large as possible. min , S max can be set by defining the lower threshold S min is set to a value that is immediately above the lower mean radiation intensity I b centered Poisson distribution, e.g. S min = I b + 2 σ(I b ), and the upper threshold S min is set to a value that is immediately below the value corresponding to the higher average radiation intensity I fcentered Poisson distribution. e.g. S max = I f - 2 σ(I f ).

[0045] Alternatively, the upper and lower threshold values ​​S can be used. min , S max also be set on the same uniform threshold value S = Smin = Smax lying between the two mean values.

[0046] Here too, the position of this value is preferably optimally adapted to the respective application. For example, in a level monitoring system used as an overfill protection device, a uniform threshold value S can be set, which is directly below the value increased by the average radiation intensity I. f lies within a centered Poisson distribution. For example, S = I f - 2 σ(I f In this way, the limit switch is able to detect and indicate a departure from the desired free state extremely early.

[0047] Accordingly, in a level monitoring system used as idle protection – as is the case in Fig. 1 is shown - preferably a uniform threshold S is defined, which is immediately above the lower mean radiation intensity I. b centered Poisson distribution, e.g. S = I b - 2 σ(I b In this way, the limit switch is also able to detect and indicate a departure from the covered state desired in this application extremely early.

[0048] The limit switch now automatically switches to operation. During operation, radiation intensities I are successively measured, and based on the radiation intensities I measured during operation and the threshold values ​​S min , S max The condition of the radiation path is monitored.

[0049] The accuracy with which the mean intensity measurements expected in the free and overcast conditions I f , Ib , and also the threshold values ​​S min , S max or S, can be determined, depends significantly on the number of individual intensity measurements I(t) recorded in the distribution V(T) of the two Poisson distributions. i ) in relation to the range of possible intensity measurements, and can be determined by the level of the minimum frequency # min to be specified. The higher the minimum frequency, the higher the achievable accuracy. min is scheduled.

[0050] A higher minimum frequency # min However, this extends the duration of the interval T. Therefore, in order to start operations as quickly as possible, a relatively low minimum frequency can initially be deliberately chosen. minPredefined threshold values ​​are used to determine preliminary threshold values ​​relatively quickly, allowing the limit switch to begin operation. In this case, a further interval TS is preferably connected to interval T in parallel with the operation initiated based on the preliminary threshold values. In this interval, the distribution V(TS) of the intensity measurements I(t) is recorded. i ) is continued until based on a significantly higher minimum frequency # max the final threshold values ​​S min , S max can be determined, which can then be used for further operation.

[0051] In applications where a difference exists between the Poisson distribution, which represents the statistical distribution of the radiation intensities incident on detector 7 in the free state, and the Poisson distribution, which represents the statistical distribution of the radiation intensities incident on detector 7 in the covered state, is... Fig. 3. If a previously known exclusion zone SP, shown in hatched detail, exists which exclusively includes radiation intensities that cannot be assigned to either the free or the covered state, the limit switch can also be used during the interval T based on the radiation intensities I(t) measured in the interval T. i ) perform a level monitoring. An exceedance of the monitored level is detected when the respective intensity measurement I(t) i ) below the restricted area SP, and a fall below the limit level to be monitored is detected when the respective intensity measurement I(t i) above the restricted area SP.

[0052] In this variant, when recording the distribution V(T), preferably only those radiation intensities I(t) measured in the interval T are included. i ) included in the distribution V(T) that lie outside the restricted area SP. 1 Filling material 3 containers 5 radioactive sources 7 Detector 9 Scintillator 11 Photomultiplier 13 Impulse line 15 Measuring instrument electronics 17 counters 7 p.m. 21 storage 23 electronic unit

Claims

[1] Radiometric limit switch, with - a spotlight (5), ◯ sends radiometric radiation along a radiation path running at the height of the limit level to be monitored, or which is in a free state when the limit level has been undershot, and or which is in a state covered by a medium when the limit level is exceeded, - a detector (7) inserted at the end of the beam path, which is designed to measure discrete radiation intensities (I) arriving at it, which depend on the state of the radiation path, wherein a statistical distribution of the radiation intensities (I) in the free and in the covered state is given by two separate Poisson distributions, - a measuring instrument electronics (15) connected to the detector (7), which ◯ an interval (T) starts in which a successive distribution (V(T)) of the measured radiation intensities (I) is recorded, o the distribution (V(T)) is continuously or at predetermined time intervals examined to see if it contains two separate Poisson distributions, o the interval (T) ends after two separate Poisson distributions have been identified in the distribution (V(T)) recorded up to that point, o based on the positions of the two Poisson distributions identified in the distribution (V(T)) within the distribution (V(T)) an upper threshold (S max ) and / or a lower threshold (S min ) for the radiation intensity (I), and o those in case of exceeding or falling below the threshold value (S min , S max ) detects a change of state to the free or covered state by measuring the radiation intensity (I) following the interval (T). [2] Limit switch according to claim 1, wherein the detector (7) converts incident radiation quanta into electrical pulses (k), and the discrete radiation intensities (I) in the form of pulse rates (N(t) i )) are measured, each corresponding to the number (n) of electrical impulses (k) per unit of time (Δt). [3] Method for monitoring the level using the radiometric level switch according to claim 1 or 2, comprising the following method steps: - by means of the emitter (5) radiometric radiation is sent along the radiation path running at the height of the limit level to be monitored, which is in the free state when the limit level is undershot, and which is in the state covered by a medium when the limit level is exceeded, - by means of the detector (7) inserted at the end of the beam path, the discrete radiation intensities (I) arriving at it, which depend on the state of the radiation path, are measured, the statistical distribution of which in the free and in the covered state is given by two separate Poisson distributions, - the interval (T) is started in which both states are occupied at least once, and in which successive radiation intensities (I) are measured, - the distribution (V(T)) of the radiation intensities (I) measured in the interval (T) is recorded, - the distribution (V(T)) is continuously or at predetermined time intervals examined to see if it contains the two separate Poisson distributions, - the interval (T) is terminated after the two separate Poisson distributions have been identified within the distribution (V(T)) recorded up to that point, and - Based on the positions of the Poisson distributions identified in the distribution (V(T)), the upper threshold (S) is determined within the distribution (V(T)). max ) for the radiation intensity (I), the exceedance of which, by a radiation intensity (I) measured after the interval (T), a change of state to the free state is detected, and / or - the lower threshold (S min ) for the radiation intensity (I) is determined, at which the change of state to the clouded state is detected by the radiation intensity (I) measured after the interval (T). [4] The method of claim 3, wherein - the means of the two Poisson distributions identified in the distribution (V(T)) are determined, - based on the mean of the Poisson distribution centered around a lower mean radiation intensity in the distribution (V(T)), an expected mean radiation intensity (I) in the clouded state b ) is determined, and - based on the mean of the Poisson distribution centered around a higher mean radiation intensity in the distribution (V(T)), an expected mean radiation intensity in the free state (I) f ) is determined. [5] Method according to claim 3, wherein the positions of the Poisson distributions identified in the distribution (V(T)) are determined based on the corresponding mean of the respective identified Poisson distribution. [6] Method according to claim 3, wherein the threshold values ​​(S min , S max ) based on location and variance (σ f , σ b ) of the two Poisson distributions identified in the distribution (VT)) are determined. [7] Method according to claim 3, wherein - the interval (T) is a period of time prior to the commencement of monitoring operation by the limit switch, - the limit switch following the interval (T) the threshold values ​​(S min , S max ) determines, and - then automatically transitions to a border control operation, -- in which successive radiation intensities (I) are measured, and -- based on the measured radiation intensities (I) and the threshold values ​​(S max , S min ) the state of the radiation path is determined. [8] The method of claim 3, wherein - a previously known blocking region (SP) lies between the Poisson distribution, which represents the statistical distribution of the radiation intensities (I) incident on the detector in the free state, and the Poisson distribution, which represents the statistical distribution of the radiation intensities (I) incident on the detector in the covered state, and which includes exclusively radiation intensities (I) that cannot be assigned to either the free or the covered state, - the limit switch performs limit level monitoring during the interval (T) based on the radiation intensities (I) measured in the interval (T), in which o an exceedance of the limit level to be monitored is detected when the respective intensity measurement (I(t) i )) below the restricted area (SP), and o a fall below the limit level to be monitored is detected when the respective intensity measurement (I(t) i)) above the restricted area (SP). [9] Method according to claim 8, wherein when recording the distribution (V(T)) only those measured radiation intensities (I) which lie outside the exclusion zone (SP) are included in the distribution (V(T)).

Images