Power-constrained networked process control system design

Abstract

Networked control systems (NCSs) operate at the interface between the physical world and cyberspace enabling the realization of emergent cyber–physical systems that can be controlled over long distance, potentially even from the cloud, thereby reducing the complexity associated with wired solutions. Networked process control systems (NPCSs), as proposed herein, incorporate the challenges inherent in NCS design within the context of process control (the application of automatic control theory). For this purpose, the essential components of an NPCS are considered to be: on the process side, the presence of dominant unstable dynamics, modeled by a first-order (or at most a second-order) transfer function; and on the network side, the presence of additive white noise (AWN) channels on both over the control and feedback paths. The contribution of this work lies in the development of new optimal designs for the stabilization of a first-order unstable NPCS that minimize the associated channel power constraints, thereby mitigating signal distortion, and preserving battery life in cases where the transmitters are not connected to an external power supply. The NPCS optimal controller configures the closed loop dynamics to exhibit a repeated stable pole, while simultaneously minimizing one of the following objectives: the input power of the AWN channel over the control path; the input power of the AWN channel over the feedback path; or the sum of the input powers of both AWN channels, all under a constant reference signal. To illustrate the theoretical developments, a first-principles-based linearized ball and beam model is derived, demonstrating the effectiveness of the proposed NPCS optimal controller designs, and their properties, in comparison to those of a classical proportional-integral (PI) design. The contributions presented in this work are readily extendable to any arbitrary linear plant model possessing a single unstable pole.