Steffen, T., Stobart, R., and Yang, Z., "Challenges and Potential of Intra-Cycle Combustion Control for Direct Injection Diesel Engines," SAE Technical Paper 2012-01-1158, 2012, doi:10.4271/2012-01-1158.
The injection timing of a Diesel internal combustion engine typically follows a prescribed sequence depending on the operating condition using open loop control. Due to advances in sensors and digital electronics it is now possible to implement closed loop control based on in cylinder pressure values. Typically this control action is slow, and it may take several cycles or at least one cycle (cycle-to-cycle control). Using high speed sensors, it becomes technically possible to measure pressure deviations and correct them within the same cycle (intra-cycle control). For example the in cylinder pressure after the pilot inject can be measured, and the timing of the main injection can be adjusted in timing and duration to compensate any deviations in pressure from the expected reference value.This level of control can significantly reduce the deviations between cycles and cylinders, and it can also improve the transient behavior of the engine. Reduced variations also mean that a better compromise between emissions and fuel efficiency can be achieved, which promises significant improvement in both.The paper looks at the technical challenges involved in the implementation of an intra-cycle control scheme. These include the fast computation speed required by the controller, the synchronization of the control algorithm to the engine rotation, and the question at which point in time (or crank angle) the control is most effective. Experimental results are used to assess the levels of pressure deviations usually seen in Diesel engines under different operating conditions, and to discuss potential control strategies to address these. The presented results demonstrate that intra-cycle control is technically feasible, and that it has the potential to reduce pressure variations and to improve both emission and fuel economy. The results also help to understand the nature of the control problem, how to measure and predict the system behavior, and the control authority that can be exercised.