Radiation measuring apparatus and method.

摘要

This invention relates to apparatus for and a method of measuring radiation field strength which uses a pulse enabled detector (Geiger-Mueller tube) and is based on the equation R = K/t, where R is the radiation field strength, t is the time till first strike, and K is a proportionality constant for the given apparatus. The apparatus and method substantially eliminate measurement errors due to circuit propagation time delays and slow drift conditions, Geiger-Mueller tube turn-on time and response curve variations, and the need to manually calibrate the measurement apparatus in high radiation fields to compensate for time errors. When mean times are short the calculation of the rate is an estimation of the mean time to strike, which in turn is equivalent to estimating the mean (1/nR) of an exponential probility distribution function (PDF) nRe<-nRt>. This function has the property that the expected value for the arithmetic mean of any right hand segment of the curve is always the same provided that the time axis is adjusted so that the starting point of the selected segment is zero. The curve is trimmed dynamically each time a new rate is calculated.
<??>An individual count of strike events is kept (28) for N separate but consecutive time units, and the counts for the later occurring events are lumped together as the Main Event. After the data collection period ends, a total accumulated time and accumulated count are constructed. This construction begins by examining the Main Event data. If enough data is present to derive a statistically acceptable estimate (250-500 events) the data in the first N time units is ignored. If more data is desirable, the data is factored in, one time unit at a time starting backward from the event count corresponding to the time unit with the longest time. The total time is now adjusted so that the starting time corresponding to the selected portion of the PDF is zero. This adjustment also includes a further adjustment which compensates for the digital approximation of the continuous variable time, corrects most of the error caused by the discrete clock frequency, and helps to adjust for additional curvature errors which are most pronounced when the mean time to strike is close to the clock time. An estimate for rate is made by dividing the total adjusted number of counts by the total adjustment accumulated time and multiplying by a scale factor.