Both European parliament and Council have defined he engine misfiring, on the 98/69/EC directive, as follows: "lack of combustion in the cylinder of a positive ignition engine due to the spark absence, poor fuel metering, poor compressione or any other cause. Interms of OBD monitoring this means that the percentage of misfires out of a total number of firing events (as declared by the manufacturer), who would result from emissions exceeding the limits or from percentage that could lead to an exaust catalyst, or catalyst overheating, causing irreversible damage." The normative defines the EOBD constraints, but a good misfire stragety must be also subject to technical and production constraints. Misfiring detection and identification in a 8 cylinder engine with a combustion every 90° of crankshaft is a particularly challenging task, given the combustion phase intersection between different cylinders. Different algorithms have been proposed in literature to reach the task. Among them, a particular mention must be placed on a strategy proposed by Azzoni et al., which is based on a mesurement of crankshaft angular speed. In Magneti Marelli Engine Control Research Department, three different strategies were conceived, implemented tested and compared: 1) A strategy based on the mentioned Fourier analysis of the crankshaft speed fluctuations; 2) A strategy based on Neural networks elaboration of the teeth period; 3) An empirical strategy (called MEDOC strategy) which compares the crankshaft acceleration with the values given by the nearest TDC's (Top Dead Center). The last strategy is deeply investigated in this paper, and will be shown how the algorithm can guarantee performances comparable to those requested by the EOBD norm constraints. The purpose of the paper is also to look at software implementation simplicty and calibrations tuning quickness. These aspects are essential in a commercial factory environment, where time to market and coste effectiveness are crucial paprameters in its economy.
Misfiring Detection in 8 Cylinders Turbo Engine
Ruggeri M;
2001
Abstract
Both European parliament and Council have defined he engine misfiring, on the 98/69/EC directive, as follows: "lack of combustion in the cylinder of a positive ignition engine due to the spark absence, poor fuel metering, poor compressione or any other cause. Interms of OBD monitoring this means that the percentage of misfires out of a total number of firing events (as declared by the manufacturer), who would result from emissions exceeding the limits or from percentage that could lead to an exaust catalyst, or catalyst overheating, causing irreversible damage." The normative defines the EOBD constraints, but a good misfire stragety must be also subject to technical and production constraints. Misfiring detection and identification in a 8 cylinder engine with a combustion every 90° of crankshaft is a particularly challenging task, given the combustion phase intersection between different cylinders. Different algorithms have been proposed in literature to reach the task. Among them, a particular mention must be placed on a strategy proposed by Azzoni et al., which is based on a mesurement of crankshaft angular speed. In Magneti Marelli Engine Control Research Department, three different strategies were conceived, implemented tested and compared: 1) A strategy based on the mentioned Fourier analysis of the crankshaft speed fluctuations; 2) A strategy based on Neural networks elaboration of the teeth period; 3) An empirical strategy (called MEDOC strategy) which compares the crankshaft acceleration with the values given by the nearest TDC's (Top Dead Center). The last strategy is deeply investigated in this paper, and will be shown how the algorithm can guarantee performances comparable to those requested by the EOBD norm constraints. The purpose of the paper is also to look at software implementation simplicty and calibrations tuning quickness. These aspects are essential in a commercial factory environment, where time to market and coste effectiveness are crucial paprameters in its economy.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
