Document flow adaptive management system with using of multilayer neural networks for aircraft production enterprise

Abstract

Document flow adaptive management system with using of multilayer neural networks for aircraft production enterprise

Savchenko E.S.

Incoming article date: 16.08.2015

In the paper we present the design and construction approach of adaptive document management system (ADMS) for the aircraft manufacturing enterprise. We propose a procedure for document workflow simulation providing adaptation to changing of life cycle of the project. For simulation of section describing study of document we use a lot of mathematical modeling tools for document workflow simulation supplemented with neural networks, which is used as an adaptive controller. The purpose of the control is achieved by minimizing the adaptation algorithm by method of error back propagation. The block diagram of ADMS nonlinear dynamic object with reference model is presented. For the purpose of nonlinear management system design, document flow nonlinear object was presented as a universal equation of Van der Pol–Relay oscillator. We explore the possibility of the document implementation that confirms the theoretical principles of the method of analytical design of aggregated regulators (ADAR). The proposed approach allows improving specialized document management system for different enterprises, particularly in the aircraft industry.

Keywords: Document workflow, adaptive system, neural network, method of synthesis, modeling