Artificial intelligence assisted MOPSO strategy for discerning the exergy efficiency potential of a methanol induced RCCI endeavour through GA coupled multi-attribute decision making approach

Abstract

The present study undertakes a comprehensive effort to explore the exergy efficiency-emissions-stability-combustion quality characteristics of premixed methanol with diesel reactivity controlled operation. The combustion phasing of the dual fuel operation primarily depends on the methanol participation rate, wherein the peak in-cylinder pressure and heat release rate decreases with increased methanol injection duration. Though simultaneous reductions of NOx and soot were observed in this study, the stability of the operations deteriorates along with unburned hydrocarbon (UHC) and carbon monoxide (CO) with increasing methanol participations resulting in a trade-off situation. Besides, the severe instability of the operation at 50% load causes misfire due to excessive dilution of the charge at higher methanol participation. The study further explores the potential of Gene Expression Programming assisted meta-model coupled Multi-objective Particle Swarm optimization (MOPSO) algorithm based multi-objective optimization endeavour to explore the optimal operational design space considering the multiple responses of exergy efficiency, NOx, PM, UHC, CO, Coefficient of Variance of indicated mean effective pressure (COVIMEP), lowest normalized value (LNV) of indicated mean effective pressure. In this present case of study, the optimization endeavor has yielded 350 numbers of Pareto solutions, while only 26 numbers of Pareto solutions were observed in the experimental counterpart. Moreover, the experimental domain of the present study has produced only single set of experiment which can satisfy the respective emission limits of NHC and PM altogether, whereas 13 sets were evident in the optimization study to maintain the NHC and PM footprint under the emission constraints simultaneously. The overall analysis of the Pareto solutions evolved in the optimization study has revealed that to attain the minimum NHC and PM footprints, the penalty of exergy efficiency and CO emissions must be incurred. The overall minimum of NHC footprint in the optimization study was recorded as 4.19 g/kWhr, compared to 6.58 g/kWhr as observed in the experimental endeavor. Similarly, the footprint of minimum PM was discovered as 0.13 g/kWhr in the optimization regime, which was 27% lower than the experimental counterpart. However, an imperceptible penalty of 1% was incurred in exergy efficiency, despite of significant lowering of overall minimum CO emissions by 84% in the optimization endeavor compared to the experimental counterpart.