<div class="section abstract"><div class="htmlview paragraph">A novel algorithm-based approach is employed in this publication to calculate multiple direct injection patterns for spark ignition engines. The algorithm is verified by investigating the combustion and emission behavior of a single-cylinder research engine. State-of-the-art standard exhaust gas analyzers, a particle counter and an additional FTIR analyzer enable in-depth investigation of engine exhaust gas composition.</div><div class="htmlview paragraph">With the upcoming worldwide pollutant emission targets, the emission limits will be reduced while the test procedures’ requirements to the engine increase. Special attention to the engine-out emissions must be paid during cold-start, during which the aftertreatment system lacks sufficient pollutant emission conversion efficiency.</div><div class="htmlview paragraph">With advanced injection control, the engine-out emissions can be reduced and exhaust aftertreatment heat-up can be accelerated. Such injection strategies separate the injected fuel quantity over several injection events for different purposes, respectively. However, increasing effort for engine calibration prevents the full employment of the advantages of complex, multiple injection strategies.</div><div class="htmlview paragraph">The algorithm-based approach used in this publication facilitates automatic injection pattern generation, instead of calibrating the multiple injections based on conventional Design of Experiment methods that are associated with extensive engine testing. To adapt to different operation conditions, a reduced number of calibration parameters are introduced to efficiently identify beneficial injection patterns. Special focus is placed on optimizing the engine cold-start behavior in view on engine-out emissions and rapid exhaust heating. Consequently, investigations have been undertaken with the research engine cooled down to -7°C. Combustion imaging using an endoscopic optical access complement to the thermodynamic measurement data and visualize the influence of injection strategy on combustion. The presented results proof that the automatically calculated injection patterns facilitate combustion optimization.</div></div>
