Design and Implementation of a Fuzzy Logic Algorithm Embedded in a Microcontroller for Combustion Applications
摘要
The present work proposes a simple real-time implementation of a fuzzy logic algorithm to control a combustion process. Fuzzy logic control’s advantage over traditional controllers is that it interprets systems as black boxes. This difference makes the control algorithms less expensive since the fuzzy logic controller does not need a high-capacity processor or precise measuring instruments. The fuzzy logic algorithm is implemented to control the regulation of the mass flow (fuel) by sensing the presence of temperature and radicals in a combustion process. The combustion process is a premixed flame produced with a commercial gas head torch fed with Liquefied Petroleum Gas at different equivalence ratios. The simplicity of the real-time implementation is in the way of applying the pointer array scheme for rule evaluation. This is confirmed by the controller response speed of 20 ms settling time regarding response speed regulating the mass flow (fuel) by sensing the presence of radicals. The proposed algorithm’s effectiveness and performance are compared with the Mamdani and Takagi–Sugeno–Kang (TSK) fuzzy inference systems using the MATLAB fuzzy logic toolbox. Therefore, using pointer-type variables in rule evaluation substantially reduces computational complexity. TSK performs better in implementing ratio control for a combustion process, although obtaining the linear functions is more complicated.