Abstract
Today light duty engines are equipped with an increasing set of embedded control systems.
The high number of software algorithms to be developed can be managed only with new robust and multi-objective engine calibration methods and tools, which potentially allow more flexibility and accuracy with respect to the traditional calibration practice. A general purpose model-based calibration methodology will be used, here named ‘the Multi-Map Optimization’ (MMO), which merges statistical concepts, like the robust design theory of experiments, numerical optimization techniques and algorithm simulation and modelling of Engine Control Unit (ECU).
In the present work this methodology has been integrated in the control algorithm design process, thus speeding up, simplifying and improving the whole process. The multi-map optimization has been used to individuate and validate the algorithm that best fits the spark advance calculation for the ECU of a gasoline engine equipped with the Variable Valve Actuation (VVA). The optimum spark advance, usually depending mainly on engine speed and air inlet pressure, depends, on VVA engine, also on VVA actuation angle. The availability of this specific optimization tool has speeded up the design of new algorithm, and, subsequently, its calibration, increasing the achieved accuracy.
Keywords: algorithm, spark, advance, valve, actuation