Multivariable optimisation of a Homogeneous Charge Microwave Ignition system

This paper attempts to address fundamental issues facing Internal Combustion Engines (ICEs) such as relatively low energy efficiency and high exhaust emissions. In particular, electromagnetic optimisation for Homogeneous Charge Microwave Ignition (HCMI) system is studied, aiming to combine the advantages of a Spark Ignition with those of a Compression Ignition system. Computational simulations of a HCMI system are carried out with three dimensional results of multi-variate changes, where the high computational load precludes a conventional iterative Computer-Aided Design process. Electromagnetic fields inside the combustion changer of the ICE are therefore optimised with a number of candidate antenna designs using various optimisation algorithms, including the Genetic Algorithm and the Nelder-Mead search algorithm. Interfaced with the Finite Element simulation software COMSOL that is used to model the engine cavity, these two a posteriori optimisation techniques are shown to be able to optimise specific system designs, with merits and drawbacks of the individual optimisation methods compared.