Using Optimization Algorithms for Control of Multiple Output DC-DC Converters

TL;DR

This paper introduces optimization algorithms to accurately determine weighting factors for controlling multiple output DC-DC converters, significantly enhancing regulation performance and transient response.

Contribution

It proposes the use of ICA, PSO, and ACO algorithms for optimal weighting factor estimation, with ICA showing superior speed and accuracy.

Findings

ICA outperforms PSO and ACO in speed and accuracy

The proposed method significantly improves output regulation

Fuzzy Logic Controller enhances transient response

Abstract

The weighted voltage mode control represents a method for control of multiple outputs DC-DC converters. Accordingly, the weighted control redistributes the error among the outputs of these converters, and the regulation error can be reduced by adjusting the weighting factors. But the problem is that most designs are performed on the trial-and-error basis, and the results were rather inconsistent. Also, in conventional mathematical approaches, this factor is designed for converters by given parameters. In this paper, three optimization algorithms namely Imperialist Competitive Algorithm (ICA), Particle Swarm Optimization (PSO) and Ant Colony Optimization (ACO) are proposed for a quick and accurate estimation of the optimal weighting factors and improve the amount of regulation on outputs of multiple outputs forward DC-DC converters. Furthermore, Fuzzy Logic Controller (FLC) is utilized to minimize the total steady-state error and improve transient characteristics by controlling the duty cycle of the PWM controller. Simulations have been performed in several cases and results show that the proposed method improves the outputs cross regulations in multiple outputs forward DC-DC converters significantly. ICA based weighting factor estimator has higher speed and accuracy in comparison with two other presented algorithms (ACO and PSO) and so is more effective in comparison with them.