Multi-Region Statistical Modeling for Dual CVVT Engine

摘要

Statistical model-based method was used to represent the characteristics of dual CVVT engine. Entire operable range was divided into three regions: stoichiometric region, WOT region and high speed part load region with enrichment. To obtain more accurate models, two-stage modeling approach was selected. Spark timing was used as local variable and other variables except spark timing were used as global variable. Test points corresponding to the combination of global variables were selected by DoE method and spark timing sweep was done on each test points. The regions impossible to operate due to combustion instability and component protection were removed before DoE using preliminary experiments. After measuring torque, emissions, combustion stability etc., statistical model building followed. Model validation was done using effective statistics and other dataset except modeling data.