Design of random inspection rate for a flexible assembly system: A heuristic genetic algorithm approach

Because of parts failure in a flexible assembly system (FAS), in-process quality control is necessary. To maximize profit, it is important to derive the inspection rate for each cell. This paper proposes an analytical model using an open queuing network theory for an FAS with feedback and reworking. The problem is solved by a modified genetic algorithm which has two major characteristics. First, the probability of mutation is variable rather than fixed in processing the search. Second, a specified mutation operator with calculating decoded values is proposed according to the status of the solution. These modifications enhance the efficiency and accuracy of the algorithm in empirical experiments. (C) 1998 Elsevier Science Ltd. All rights reserved.