Maximum-Entropy Principle in Flexible Manufacturing Systems

Abstract

It is shown that the entropy of the joint probability distribution of the queue lengths of the M machine-groups in a closed queuing network model of a flexible manufacturing system is maximum when the loads on the different machine-groups are equal for both single-machine machine-groups and for those multiple-machine machine-groups of which group sizes are equal. It is also shown that for unequal machine-group-sizes, the entropy is not maximum when the workload is balanced. The simultaneous variations of the entropy function of the load distribution, the entropy function of joint probability distribution lengths of queues and the expected production function are studied in order to investigate the relationship between the information content and productive capacity of manufacturing systems. Four measures of load balance in a flexible manufacturing system are given.