Neural networks implementations to control real-time manufacturing systems

The main objective of advanced manufacturing control techniques is to provi de efficient and accurate tools in order to control machines and manufactur ing systems in real-time operations. Recent developments and implementation s of expert systems and neural networks support this objective. This resear ch explores the use of neural networks to control several manufacturing sys tems in real-time operations: robot manipulators, tool changes, conveyor sy stems and machine faults diagnosis. The main barrier to wide implementation of neural networks is the huge computation resources (times and capacities ) required to train a network. This research represents the use of a multi- layer architecture of networks (input layer, several hidden layers and an o utput layer) to define single-valued inter-relationships between system par ticipants and to avoid the need for long training processes. The use of neu ral networks to control the above-mentioned systems was evaluated from the following parameters: the architectures, network training methods, efficien cies and accuracies of networks to perform the task of control, Several con clusions related to neural network implementations to manufacturing systems were produced: (1) the multi-layer architecture fits the complexity of man ufacturing systems; (2) neural networks are efficient to control real-time operations of machines; (3) machines which were controlled by neural networ ks performed accurate results; and (4) the use of several hidden layers can replace the need for long training processes and saves on computation reso urces. (C) 1998 Published by Elsevier Science Ltd. All rights reserved.