Current Statistical Quality Control of Technological Process

Victor Aleksandrovich Vlasov¹, Svetlana Victorovna Vlasova² and Tolokonsky Andrey Olegovich¹

¹National Research Nuclear University «MEPhI» (Moscow Engineering Physics Institute), Russian Federation, 115409, Moscow, Kashirskoye, 31. ²All-Russian Research Institute for Nuclear Power Plants, Operation,Russian Federation, 109507, Moscow, Ferganskaya, st., 25

ABSTRACT: The task of statistical testing of changing quality of technological process in case of making changes in the process for the purpose of quality improvement is considered in the article. Almost every technological process is subjected to random disturbances which mean that the current quality of technological process is casual. It is described by the special parameter connected with the distribution law of random variables xi the values of which after the experiment are the certain values ci. Such parameters, for example, could be a failure rate l of products in large let-outs, an average sports result of an athlete, an average life expectancy of a patient after considerable treatment (operational interventions etc.). The greatest difficulty of statistical control is that it is necessary to face the procedure of testing composite parametrical hypotheses. For example, the hypothesis H can be a statement that the failure rate is less than the predetermined value (l < l0, where l0 - is a predetermined value). Usually for testing composite parametrical hypothesis the distribution law of unknown parameter is applied. However this law in the majority of practical situations is unknown, and this fact complicates the procedure of correct reduction.

KEYWORDS: Control; technological process; simple; Composite parametrical hypotheses; substitution for reduction

[ HTML Full Text]