The Scheduling of Flexible Manufacturing System Based on RBF Neural Network

Improving the procedure management for flexible manufacturing system (FMS) is still one of the main topics in present industries. In this paper, a radial basis function neural network (RBF-NN) model is proposed to schedule jobs in a general FMS. First, a FMS and the problem are described. Then, a RBF-NN is presented; the iterative training algorithm employs the gradient algorithm to minimize a function that measures the difference between the network output and the desired one. Finally, according to some scheduling rules the ideal configuration of the RBF-NN that is used for the criterion of mean tardiness is figured out. The ideal configuration of the RBF-NN has 8 input nodes, 17 nodes in the hidden layer, and 13 nodes in the output layer. Simulation results show that the combinations are the optimal strategies.