Depth of planning the state of a dynamic discrete system by autocorrelation function

TL;DR

This paper models a multidimensional production system as a dynamic discrete system and uses autocorrelation functions to analyze its state and cyclical behavior, enabling advanced analytical methods.

Contribution

It formalizes the production system with a large parameter set and demonstrates the use of autocorrelation functions for state analysis and system dynamics understanding.

Findings

Autocorrelation functions reveal cyclical dynamics in the production system.

Modeling with 1.2 million parameters enables detailed system analysis.

Formalization facilitates application of advanced analytical methods.

Abstract

The production system (multidimensional object) is considered as a dynamic system with discrete time. Formalized: space (state of the object, control actions, goals, observed values, analytical estimates). Analytical estimates of the state of a dynamic system are formed through the autocorrelation function. The autocorrelation function is calculated with the regulator setting the length of the analyzed time series (analysis depth). A digital copy of the production system is created, characterized by 1.2 million parameters. Modeling the activities of the production system is performed in the author's complex of programs. In total, twenty-eight controller states are calculated to analyze the effect of repeating parameters affecting the activity of the production system. The simulation shows the cyclical dynamics of changes in the autocorrelation function. Formalization of the production system is carried out, which allows you to move on to other methods of analysis of the production system: Kalman filter, neural network forecast, recurrence equation, balances.