Automatic voltage regulator design based on multi-objective optimisation concept

In this paper, a novel analytical two degrees of freedom (2DOF) controller for the automatic voltage regulator (AVR) system is proposed, with an explicit tuning formula derived based on a desired appropriate disturbance transfer function to enable precise control of the closed-loop performance and sensitivities. Unlike the conventional formulation of AVR control as a constrained single objective optimisation problem, the proposed approach optimises the controller through the multi-objective particle swarm optimisation (MOPSO) algorithm, using the best-desired disturbance transfer function. Three conflicting control objectives are considered, represented by disturbance integral absolute error (IAEd), maximum sensitivity (MS) and maximum noise sensitivity (MN). The resulting Pareto optimal sets clarify, for the first time, the trade-off on the AVR controller design. A comparative analysis with the most recent 2 DOF controllers has been performed where the role of the controller structure appears clearly in both performance and robustness. A new approach to AVR controller optimisation is introduced in this work, emphasising a prioritised balance between performance and robustness while leveraging analytical tuning.