Neural Network Decoder of Automatic Process Control System

In this paper the new associative neural network calculating Lee code distance which allows defining in real time the closest pattern in residue number system is considered. The presented neural network allows to detect and correct the errors in data transfer in automatic control systems with high error rate. The neural network able to form the limited list of the most probable sets of residues from legitimate range of residue number system representation in case of insufficient redundancy for error correction. Designed neural network consist of simple modulo adders and lookup tables, and is able to convert data stream from residue number system to positional number system with error correct.