Distributed Computational Architecture for Industrial Motion Control and PHM Implementation

Abstract

Computational architecture is a major challenge in implementing "Prognostic Health Management (PHM)" solutions, for many industrial applications. Specially for "Industry 4.0" requirements, the computational architectures should be evolving as per computational requirement, computational power and communication capability within the involved devices. This work proposes a distributed computational architecture and its utilization in industrial application. The distributed control development and its usage as edge intelligence for industrial applications, has been discussed. A novel neural network architecture is proposed, which could be used for integrating industrial domain knowledge with machine learning technique, in the context of PHM implementation.