Title: Multi-objective optimisation of injector and diesel engine by genetic algorithm: Nu-SVR modelling

Authors: Hadi Taghavifar; Simin Anvari

Addresses: Department of Mechanical Engineering, Faculty of Engineering, Malayer University, Malayer, Iran ' School of Chemical Engineering and Technology, Xi'an Jiaotong University, Xi'an, China

Abstract: The current study deals with application of evolutionary multiobjective genetic algorithm (MOGA) in a 1.8 L Ford diesel engine to enhance the power, fuel consumption, and air-fuel uniformity. To do so, four design parameters of engine geometry and injector parameters are defined and three sub-objectives are considered to get either maximised or minimised. On the sub-objectives, constraints are imposed to introduce the feasible solutions. The best solution is obtained at RunID66 from 70 design points. The results showed that increasing bowl radius after a certain point is not useful for enhancement of the mixture homogeneity. In the second part of the study, support vector regression (SVR) technique is applied on the input and output data to make a model to predict the engine power.

Keywords: diesel engine; diesel injector; Nu-SVR; MOGA; multi-objective genetic algorithm; Pareto front.

DOI: 10.1504/IJHVS.2020.108732

International Journal of Heavy Vehicle Systems, 2020 Vol.27 No.3, pp.340 - 358

Received: 21 Feb 2017
Accepted: 24 Jul 2017

Published online: 30 Jul 2020 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article