Calibration of a type 2 controller applied to a buck converter: a multi-objective analysis

Abstract

Control lops are nowadays everywhere, from tiny devices to robust industrial applications. However, even when only few parameters are being dealt, manually fine-tuning has been shown to be a meticulous task for achieving designers' desired performance. Fine-tuning a controller parameters is an arduous job that requests expertise of the domain. On the other hand, metaheuristics have been barely applied for accomplishing this task. In particular, due to different behaviors that a control loop can have, a multi-objective analysis shows up as essential. In this work, we apply a Multi-Objective Variable Neighborhood Search based algorithm for assisting the design of a Buck Converter, integrating the optimization process with an evaluation mechanism integrated with a circuit simulation software. The obtained Pareto Front presented various response behaviors, optimizing different desired characteristics. We suggest that the proposed framework is a promising tool for assisting decision makers to design more efficient and dynamic systems.