Evaluation of Statistical Reliability Measures for a Program for the Limitation of Power Density in a Reactor Vessel

Abstract Errors within an assembler program for limiting the local power density in a reactor vessel are to be identified and corrected during two test phases. According to the special features for error detection of the testing facilities, the errors are classified, and an appropriate software reliability model is applied to each error class. These models allow prediction of certain reliability parameters from the observation of the test. Depending on the model employed these parameters can be e.g. the number of remaining errors, the probability for the occurance of an error within a given time interval, the meantime between successive errors. In this contribution references are only made to the number of remaining errors.