Statistical assessment of the reliability of decisions on the state of controlled technological process based on E. Volodarskyi approach

The article concerns the issues of using control charts to study the parameters describing the state of the productsin their production process. The identification of reliability of a decision is based on the assessment of disturbances occurring in it is discussed.Using is the method proposed by † Yevhen Volodarskyi1, that was based on Bayesian approach. The influence of measurement errors and their distribution of probability on the correctness of the decisions taken is considered. In the article two estimations of conformity of technological process to the norms based on the results of its control are considered. The first assessment is a-priori probability or reliability of the control result, which is performed before the control procedure and is based on a-priori data about the process and measurement error. The paper proposes the use of the second assessment, namely the posterior probability of compliance of the technological process with the norms. This assessment of compliance is performed after the control result is obtained, when only half of the set of elementary events contributing to the occurrence of one of the control results is left for evaluation. The use of this estimation allows doubling the statistical reliability of the control result estimation. The effectiveness of assessing the compliance of the technological process with the standards established for the uniform distribution of the values of its controlled parameters and their measurement errors is also determined.