Cyclic dynamics of misfires and partial burns in a dilute spark-ignition engine

Abstract

This study investigates the cyclic dynamics of cumulative heat release data past the edge of stability in a dilute spark-ignition engine. Emphasis is placed on analyzing the cyclic dynamics near the dilute limit where partial burns and misfires are frequent. These events are often followed by a higher-energy cycle due to the feed-forward mechanism present in the residual gases. These patterns are deterministic and increase the coefficient of variation to undesirable levels. Symbol sequence analysis was used to investigate the cyclic dynamics of these low–high patterns. The heat release was partitioned on an energy basis to give physical meaning to each partition and each sequence created when analyzing the symbol sequence results. This partitioning method provided insight into the differences in the dynamics when operating in the misfire or partial burn regime. These differences could impact the control method used.

Investigation of the cyclic dynamics in the misfire and partial burn regime in a dilute spark-ignition engine and the role of the residual

“…Data was partitioned using the energy-basis method that was outlined in Stiffler et al 11 to allow for different cutoffs to be compared directly, but with the binning cutoffs being those shown in Table 2. Cycles with nominal heat release had a combustion efficiency around 80%–90% since heat transfer and crevice effects were neglected in the heat release calculation.…”

“…The process of symbol sequencing has been discussed earlier in this paper and full details and uses in engine studies can be found in the literature. 11,13–15 Symbol sequencing only uses frequency to compare different sequences, the benefit of using additional measures available when using association rules is discussed in this section.…”

“…The process of symbol sequencing has been discussed earlier in this paper and full details and uses in engine studies can be found in the literature. 11,[13][14][15] Symbol sequencing only uses frequency to compare different sequences, the benefit of using additional measures available when using association rules is discussed in this section. When restricting the association rules where X must contain heat release from consecutive previous cycles when Y is the heat release of cycle (i), then the support values for the rules obtained are the same results obtained from symbol sequencing with a sequence length equal to the length of the entire rule.…”

“…The cycle-to-cycle dynamics at highly dilute conditions have been studied for decades, but they often do not look at long-term dynamics or do not investigate specifically how misfires and partial burns behave between the two regimes. Stiffler et al 11 found the dynamics in a highly dilute spark ignition engine (minimum COV of IMEP was 27.33%) were different in combustion efficiency (defines how much of the fuel supplied is burned during an engine cycle) were observed between operation with excessive misfires compared to operation with excessive partial burns. The authors found repetitive partial burns appeared stochastic, but were unable to conclude if differences observed between lean and EGR operation were from differences in the level of dilution and/or the composition.…”

Reinforcement learning applied to dilute combustion control for increased fuel efficiency

“…44 Moreover, the fundamental characteristics of combustion CCV can change depending on engine speed, load, and phasing. [45][46][47] Since SI engines operate over a wide range of speeds and loads, complete characterization of the dilute limit over the entire operating regime requires a large amount of time and effort. To reduce the testing burden, the Advanced Combustion and Emission Control (ACEC) tech team of the United States Council for Automotive Research (USCAR) has chosen a relevant engine speed and load condition for different technology applications.…”