Iterative learning approach for diesel combustion control using injection rate shaping

State of the art in combustion control consists of discrete and clearly separated multiple injection events. Future available injector types establish highly flexible injection, allowing for continuously variable injection timings and therewith shaping of the fuel injection rate. This leads to a more direct influence on combustion, e.g. control of time-continuous pressure traces instead of mean value control. The overall advantage of rate-shaping approaches is higher engine efficiency and reduction of engine-out emissions at the same time. However, well investigated and established model-based control approaches are not applicable anymore. For that reason, this paper presents a simple and effective control method which uses iterative learning control (ILC) combined with variable injection rate. The presented control algorithm is validated for diesel combustion. To make ILC applicable to combustion processes, the underlying learning rule will be extended by model-based combustion knowledge. Static and dynamic simulation results are presented which emphasize the relevance of the approach.