Diesel Engine performance improvement in a 1-D engine model using Particle Swarm Optimization

Nonlinear engineering: Modeling and application

article

Prashanth Karra¹ 

[1] RMIT University - School of Aerospace, Mechanical and Manufacturing Engineering 264 Plenty Road Bundoora Victoria 3082

关键词: GT-Power;    1-D simulations;    Particle Swarm Optimization;    Diesel emissions;    Tier 3 emissions;   

DOI  :  10.1515/nleng-2015-0017

来源: De Gruyter

【 摘 要 】

A particle swarm optimization (PSO) technique was implemented to improve the engine development and optimization process to simultaneously reduce emissions and improve the fuel efficiency. The optimization was performed on a 4-stroke 4-cylinder GT-Power based 1-D diesel engine model. To achieve the multi-objective optimization, a merit function was defined which included the parameters to be optimized: Nitrogen Oxides (NOx), Nonmethyl hydro carbons (NMHC), Carbon Monoxide (CO), Brake Specific Fuel Consumption (BSFC). EPA Tier 3 emissions standards for non-road diesel engines between 37 and 75 kW of output were chosen as targets for the optimization. The combustion parameters analyzed in this study include: Start of main Injection, Start of Pilot Injection, Pilot fuel quantity, Swirl, and Tumble. The PSO was found to be very effective in quickly arriving at a solution that met the target criteria as defined in the merit function. The optimization took around 40-50 runs to find the most favourable engine operating condition under the constraints specified in the optimization. In a favourable case with a high merit function values, the NOx+NMHC and CO values were reduced to as low as 2.9 and 0.014 g/kWh, respectively. The operating conditions at this point were: 10 ATDC Main SOI, -25 ATDC Pilot SOI, 0.25 mg of pilot fuel, 0.45 Swirl and 0.85 tumble. These results indicate that late main injections preceded by a close, small pilot injection are most favourable conditions at the operating condition tested.

【 授权许可】

Unknown