The method described in this paper addresses the problem of interaction in a control system that consists of multiple legacy PI(D) controllers. In these systems, it is often not possible to intercept the signals between the controllers and plant and replacement with multivariable alternatives such as model predictive control (MPC) may not be viable from a cost perspective. To address these issues, we propose a decoupling method that involves first changing the tuning of the controllers and then dynamically adjusting the setpoints. The method enables decoupling of a multiloop control system via setpoint adjustment which can be carried out without having to re-engineer any of the legacy controllers. The paper focuses on the application area of energy systems in buildings where costs are constrained and PID multiloop configurations are prevalent. Simulation test results are presented from two systems to demonstrate the improvement in control achieved with the method.

This content is only available via PDF.
You do not currently have access to this content.