Optimization Tools Based on Metaheuristics for Performance Enhancement in a Gaussian Adaptive PID Controller

IEEE Transactions on Cybernetics
Erickson Diogo Pereira PuchtaMauricio Dos Santos Kaster

Abstract

This paper presents the proposal of using two bio-inspired metaheuristics-genetic algorithms (GAs) and particle swarm optimization (PSO)-to adjust the free coefficients of a Gaussian adaptive proportional-integral-derivative (GAPID) controller. When a specific adaptation rule is imposed to a conventional proportional-integral-derivative (PID) controller, either by means of a hyperbolic tangent function or a Gaussian function, the solution is left exposed to the function restrictions/impositions. Finding the correct proportionality between the parameters is an arduous task, which often does not have an algebraic solution. Each Gaussian function of each control action has three parameters, resulting in a total of nine parameters to be defined. This paper proposes making the parameters linked to the linear PID gains, in order to keep the GAPID the same design requirements as for the PID. Then, two metaheuristics (GA and PSO) were employed in order to find the best parameters for the GAPID. A comparison between these two strategies is presented. In this investigation, a well-known plant of a step-down dc-dc converter is used, which represents a typical second-order system, where the absence of significant nonlinearities helps focus...Continue Reading

Related Concepts

Related Feeds

CREs: Gene & Cell Therapy

Gene and cell therapy advances have shown promising outcomes for several diseases. The role of cis-regulatory elements (CREs) is crucial in the design of gene therapy vectors. Here is the latest research on CREs in gene and cell therapy.