Neural network modelling and simulation of the scheduling

Abstract

A three layer feed forward neural network was constructed and tested to analyze the scheduling process on single machine. The operating variables studied are the operation, processing time, setup time, deadline time, duedate time, priority, machine, and fabric color. These variables were used as input to the constructed neural network in order to predict the scheduling completion time as the output on a single machine. Three layer feed forward network trained with error back propagation learning rule are used. The constructed network was found to be precise in modeling the scheduling for the operating conditions studied and also, in predicting the scheduling for the new input data which are kept unaware of the trained neural network. © 2008 International Federation for Information Processing.