Accession Number : AD0250263

Title :   THE AOQL OF MULTI-LEVEL CONTINUOUS SAMPLING PLANS

Corporate Author : STANFORD UNIV CA APPLIED MATHEMATICS AND STATISTICS LABS

Personal Author(s) : ELFVING, G.

Report Date : 05 DEC 1960

Pagination or Media Count : 1

Abstract : THE MULTI-LEVEL PLANS (MLP) FOR CONTINUOUS QUALITY CONTROL ARE CONSIDERED. THE DISCUSSION AIMS TO FIND THE PAYOFF LEVEL (AOQL) FOR PROCESSES NOT ON CONTROL. ONLY PROBABILITY SAMPLING, AT A FINITE NUMBER OF LEVELS CHARACTERIZED BY DECREASING BUT OTHERWISE ARBITRARY, SAMPLING RATES, AND ARBITRARY RELAXATION NUMBERS ARE CONSIDERED. UNDER A CERTAIN RESTRICTION ON NATURE'S STRATEGIES, THE AOQL IS DERIVED FOR THE PROCEDURES MLP - 1, MLP T. AND MLP - R. THE RESULTS ARE GIVEN IN A THEOREM. NATURE'S OPTIMAL STRATEGY IS SHOWN TO BE ALWAYS DETERMINISTIC, AND TO A PROCESS IN WHICH THE SAMPLING LEVELS FOLLOW EACH OTHER CYCLICALLY IN A COMPLETELY DETERMINED WAY.

Descriptors :   *QUALITY CONTROL, *SAMPLING, PROBABILITY, STATISTICAL ANALYSIS.

Distribution Statement : APPROVED FOR PUBLIC RELEASE