Section] Section] Section] Section] Section] Section] Section] Section] Section] Section] 0 Estimating Performance Measures in Stochastic Cyclic Schedules

This paper reviews a cyclic scheduling model of repetitive manufacturing environments consisting of exible machines and multi-stage jobs. We explicitly model two system performance measures, throughput and ow time, while incorporating variability arising from uncertainty in operation processing times. We present an iterative approximation scheme to estimate the two moment behavior of our performance measures. The basic approach relies on Clark's method for computing moments of the maximum of bivariate normals. The proposed algorithm, called the Cyclic Network Clark's Algorithm, is illustrated via a simple ow shop example and subjected to extensive computational testing.