COORDINATION UNDER CONDITIONS OF COMBINED UNCERTAINTY

Abstract

In this publication, we consider the problem of coordination of industrial control systems (ICS) under combined uncertainty, which includes stochastic and fuzzy uncertainty. Uncertainty can arise from various sources, such as expert forecasts, changes in the production program over time, and statistical identification of the parameters of the mutual influence of objects. The nonlinearity of the control law adds to the complexity of the study.

The paper considers the case of a DCPS with combined fuzzy and relay control on the example of a thermal control system for a multi-zone facility. A block diagram of such a system is presented and key components, such as a fuzzy coordinator and relay controllers, are described. The adjustable parameters of the system are the heater power and the width of the hysteresis zone of the local relay controllers.

To achieve optimal control under conditions of uncertainty, the paper uses operator models and generalizing uncertainty functions. The authors propose an optimization criterion based on the standard deviation of the uncertainty functions from the set temperature parameters.

 The study was performed using the Scilab/Xcos software system, where the generalizing functions were calculated in parallel in one simulation cycle. Experimental data show that there is an optimal value of the width of the hysteresis zone for each level of random effects, which ensures minimal control uncertainty. In addition, increasing the power of the energy source reduces the uncertainty, but leads to an increase in the frequency of self-oscillations and, as a result, to a decrease in system reliability.

The paper also addresses the issue of coordination of DCPS under uncertainty, investigating the robustness of these systems and the possibility of decentralized coordination under uncertainty.

As a result, the paper provides an important contribution to the understanding of the problems of coordinating DCPS under combined uncertainty and provides useful insights into the optimal management of such systems.