<abstract xmlns="http://www.w3.org/1999/xhtml">

<p>In order to actively control combustion reaction, this study proposes an adaptive neuro-fuzzy (ANFIS) control scheme of interaction between premixed combustion reaction and acoustic flame perturbation where the flame pressure movement will be considered as model perturbation. Using the Cantera database, it is possible to investigate the mechanisms by which the combustion process interacts with acoustic, vorticity, and entropy waves. A well-stirred reactor (WSR) has been extensively used to model combustion processes in three different reaction zone regimes. We designed the control architecture to achieve an intelligent representation of the system for various operating scenarios, which was motivated by the complexity of the mathematical model that was being used. This goal is accomplished by an artificial bee colony (ABC), which uses simulated data from a mathematical model to optimize a neuro-fuzzy with less computational expense. The optimized neuro-fuzzy identifier is converted to an adaptive neural-based (ANFIS) controller optimized to control the outputs of the system. In keeping with the combustion temperature set point, the results demonstrate a remarkable attenuation of flame perturbation and acceptable combustion reaction quality (NOx emission).</p>
</abstract>

In order to actively control combustion reaction, this study proposes an adaptive neuro-fuzzy (ANFIS) control scheme of interaction between premixed combustion reaction and acoustic flame perturbation where the flame pressure movement will be considered as model perturbation. Using the Cantera database, it is possible to investigate the mechanisms by which the combustion process interacts with acoustic, vorticity, and entropy waves. A well-stirred reactor (WSR) has been extensively used to model combustion processes in three different reaction zone regimes. We designed the control architecture to achieve an intelligent representation of the system for various operating scenarios, which was motivated by the complexity of the mathematical model that was being used. This goal is accomplished by an artificial bee colony (ABC), which uses simulated data from a mathematical model to optimize a neuro-fuzzy with less computational expense. The optimized neuro-fuzzy identifier is converted to an adaptive neural-based (ANFIS) controller optimized to control the outputs of the system. In keeping with the combustion temperature set point, the results demonstrate a remarkable attenuation of flame perturbation and acceptable combustion reaction quality (NOx emission).

A Bee Colony Neuro-Fuzzy Controller to Improve Well Premixed Combustion

Strojnícky časopis - Journal of Mechanical Engineering

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 3.0 License.

{"article-title":"A Bee Colony Neuro-Fuzzy Controller to Improve Well Premixed Combustion"}

In order to actively control combustion reaction, this study proposes an adaptive neuro-fuzzy (ANFIS) control scheme of interaction between premixed combustion...

A Bee Colony Neuro-Fuzzy Controller to Improve Well Premixed Combustion

Publicado en línea: 25 may 2023

Páginas: 25 - 42

DOI: https://doi.org/10.2478/scjme-2023-0003

© 2023 Debbah Abdesselam et al., published by Sciendo

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 3.0 License.