Proposal and Optimization of a Fuzzy Logic Model for Automobile Injection Diagnosis Using Evolutionary Algorithms

This article presents results for the design of a fuel injection diagnostic strategy for motor vehicles in Ciudad Juárez, Chihuahua. The first part of this research develops a diagnosis model for the injector MD162524, implementing Fuzzy Logic. Characterization of the diagnostic model took five injector operation tests (spray pattern, injector opening cycles at 3 revolution levels, electrical resistance measurements in the injector, injector flow at a constant pressure and activation of two levels of pressure in the injector) and 3 possible outcomes of diagnostic status (good shape, repairable, useless). In the final part of the investigation, two evolutionary algorithms are used as a tool to optimize the response of the model designed, generating data sets with decreasing measurement error.