Application of Improved Artificial Bee Colony Algorithm to the Parameter Optimization of a Diesel Engine With Pilot Fuel Injections

Abstract

Modern diesel engines are charged with the difficult problem of balancing emissions and efficiency. For this work, a variant of the artificial bee colony (ABC) algorithm was applied for the first time to the experimental optimization of diesel engine combustion and emissions. In this study, the employed and onlooker bee phases were modified to balance both the exploration and exploitation of the algorithm. The improved algorithm was successfully trialed against particle swarm optimization (PSO), genetic algorithm (GA), and a recently proposed PSO-GA hybrid with three standard benchmark functions. For the engine experiments, six variables were changed throughout the optimization process, including exhaust gas recirculation (EGR) rate, intake temperature, quantity and timing of pilot fuel injections, main injection timing, and fuel pressure. Low sulfur diesel fuel was used for all the tests. In total, 65 engine runs were completed in order to reduce a five-dimensional objective function. In order to reduce nitrogen oxide (NOx) emissions while keeping particulate matter (PM) below 0.09 g/kW h, solutions call for 43% exhaust gas recirculation, with a late main fuel injection near top-dead center. Results show that early pilot injections can be used with high exhaust gas recirculation to improve the combustion process without a large nitrogen oxide penalty when main injection is timed near top-dead center. The emission reductions in this work show the improved ABC algorithm presented here to be an effective new tool in engine optimization.

Optimization of diesel fuel injection strategies through applications of cooperative particle swarm optimization and artificial bee colony algorithms

“…Similar forms of the objective function have been used in other optimization studies. 12,16,23 Both ABC and CPSO 3 were applied to the engine for testing. The flow charts for these algorithms are given in Figures 3 and 4, where both algorithms use equation ( 3) to initialize food sources in ABC or particles in PSO.…”

“…ABC algorithm was used for the first time in engine optimization on a four-cylinder diesel engine operating at medium load. 23 Food sources included EGR level, main injection timing, pilot offset, pilot fuel quantity, and intake manifold temperature. A small Colony Size and MCN were used to reduce the total number of experiments.…”

A novel machine learning-based optimization algorithm (ActivO) for accelerating simulation-driven engine design

“…Karra and Kong [6] used particle swarm optimization (PSO) and experimental techniques to find the operating conditions that minimize emissions from a multi-cylinder, turbo-charged diesel engine. Zhang et al [7] used a variant of the artificial bee colony algorithm to experimentally minimize emissions from a diesel engine.…”

Comparative of various bio‐inspired meta‐heuristic optimization algorithms in performance and emissions of diesel engine fuelled with B5 containing water and cerium oxide additive blends

“…The results showed an improvement of 27% and 60% in NOx emissions and particulate matter (PM), respectively. Recently, an artificial bee colony (ABC) was incorporated with GA to optimize the engine performance and emissions in a study conducted by Zhang et al 39 They performed the tests at different fuel injection strategies, exhaust gas recirculation (EGR) rates, and intake air temperature and they obtained up to 77% NOx emission reduction under a specific condition of EGR.…”