Most of today's industrial processes being of complex and hybrid nature (consisting of both continuous and discrete event pans), require by necessity a supervisor for control. Since the supervisor is at the upper level of a two-level hybrid system, it is required to process and make decisions on a large amount of nonhomogeneous information. In this paper we discuss the categorization of information through neural network interfaces and the issues surrounding the format of the input data to the network, along with a measure of the information content of the data eventually reaching the supervisor.
All articles published by Studies in Informatics and Control (SIC) are published in accordance with the terms and conditions of the Creative Commons Attribution License (CC BY-NC-SA 4.0).