APPLICATION OF STOCHASTIC REAL-VALUED REINFORCEMENT NEURAL NETWORKS TO BATCH-PRODUCTION RESCHEDULING

This paper details the design and application of a hybrid neural netwo rk architecture for the rescheduling problem of batch manufacture. Des ign issues include the selection of an appropriate neural network para digm, specification of the network architecture and support for multis tep prediction. Application issues include decoupling the network dime nsion from that of the problem and the definition of suitable reschedu ling operators. The ensuing hybrid network is tested against heuristic s previously identified as typically representing estimates for best a nd worst case performance within a cross-section of batch rescheduling problems.