CONTINUOUSIDENTIFICATION OFA FOUR-STROKE SIENGINE

Compact eni models often consist of a of noninear differential equationm which predictthetimedevelopment ofthe mean value of the engine state vaiabks (and perhaps some internalvariabls):such modes are sometimes calledmean value engine models. Currently a great deal of attention is focused on constructing suchoontinuous timemodels andon finding their parameters. Thispapershows, thatitis posibletoidentify an enginemodel from a linearized versionofa mean value modelfor a CFIfour-cycle spark ignition (SI) engine. Suchan approach is useful because itpreseves a phyical undertanding oftheengine throughout theidentification stae.Afterwards theidentifriation results areavailable forgeneral dynamic enginestudies The identfication techniques diussedinthispaper include casical methods (step response) as well as moden satical methods (Kalman filtering andMaxmum Likeliood estimation). These techniqe havebeenapplied toa fourcylinder SIengine. Theresults include an identfiation ofthemostimportant paraneters andtimecntantsof theengine. These are of interest for the construction of engine simulation models,for control studies andcndition monitoring applications. 1.htreduc&s Inspite of thegreatinterest wbichexists inidentifying the parametesofengi (and in particular s kignitin engines) there isverylittle work ofthis naturetoredin theliterature. This is mainly becae 1. an engine is a nonlnear systemwhichisdifficult todescribe physically and