PID Controller with Real Number Genetic Algorithms

• Autor: Tejada Muñoz, Guillermo
• Formato: artículo
• Fecha de Publicación: 2020
• Institución: Universidad Nacional Mayor de San Marcos
• Repositorio: Revista UNMSM - Industrial Data
• Lenguaje: español; inglés
• OAI Identifier: oai:ojs.csi.unmsm:article/16489
• Enlace del recurso: https://revistasinvestigacion.unmsm.edu.pe/index.php/idata/article/view/16489
• Nivel de acceso: acceso abierto
• Materia: Genetic algorithms; real number genes; PID; mean square error; Algoritmos genéticos; genes de números reales; error cuadrático medio

A genetic algorithm (GA) is an artificial intelligence technique that can be applied to any engineering specialty. In this study, the proposed GA finds optimal values for PID controller parameters Kp, Ki and Kd, which are widely used in the industry. Chromosomes, composed of Kp, Ki and Kd genes, represented by real numbers, evolve and are evaluated by the mean square error (MSE) of the desired output. In that sense, the solution is the chromosome with the lowest MSE, which produces less transient output. In addition, the GA has been encoded in MATLAB language and results have been compared with other works.