Regresar

Exhaust gas pbkp_rediction in otto engines using PME and RNA

Abstract:

This document deals with the pbkp_rediction of exhaust gases from a provoked ignition engine (MEP), through its mean effective pressure (PME), through the use of artificial neural networks (ANN). The methodology applied consists of acquiring signals from the PME, load, revolutions per minute (rpm) and the manifold absolute pressure sensor (MAP) of the engine of the armfield dynamometric bench cm11, of which it obtains a database and that through statistical analysis it is determined that to pbkp_redict polluting emissions it is necessary to apply cascade RNA, that is, first the engine load is pbkp_redicted and therefore the exhaust gases, with which a rating error less than 10e-4. For the validation of this project an experimental analysis is carried out, which consists in acquiring new engine data to check the percentage of error between the values simulated by the RNAs and the actual values, where there is an error of less than 2% and 3% for the pbkp_rediction of the load and the exhaust gases respectively.