Neural network for minimizing tricriteria objective function for machine scheduling problem

Abstract

In this paper, neural networks (NN) are implemented to manipulate a problem of machine scheduling for a single machine to minimize multicriteria objective function: maximum tardiness, maximum late work and total late work simultaneously (Tmax, Vmax, ΣVj). Also the results of applying NN are compared with some exact methods, some heuristic methods and local search methods [7,14]. The used NN is learned by pack propagation algorithm for n = 3: 500 jobs. NN technique was be found to be effective and used to select the best efficient and optimal schedule which minimizes the objective function.