Simulated Annealing Based Artificial Neural Network For Real Time Dispatching

Simulated Annealing is a meta-heuristic search technique that can be used for training a multi-layered feed forward neural network. In this study we present a simulated annealing based training routine for a neural network based prediction model that can be used for predicting cycle time through queue time restricted zones under different WIP conditions inside the zone. Once trained such a model is capable of producing real-time predictions which can be used for real-time dispatching decision making.