Design of Metallic Catalytic Converter using Pareto Optimization to Improve Engine Performance and Exhaust Emissions

Abstract

In this paper, Metallic Catalytic Converter (MCC) is installed in motorcycle exhausts to produce the minimum CO as well as to produce the optimum engine power. The results from previous research were collected and then used to predict the best MCC design using the Artificial Neural Network Multi-Objective Genetic Algorithm (ANN-MOGA). In addition, the ANN parameter tuning process was also carried out using the Taguchi method to find the initial weighting and bias that is able to provide the best and the most stable performance to predict the best MCC design. The best two sets of design solutions out of 70 sets of Pareto solutions were obtained by ANN-MOGA. Those two MCC designs are the optimum emission design and the optimum multi-objective design. The verification results show that the optimum multi-objective design tends to be superior in terms of CO emissions and engine power. In terms of CO emissions, the optimum multi-objective design gets a larger S/N ratio of -10.98, while the optimum emission design only gets an S/N ratio of -11.21. Meanwhile, in terms of engine power, the optimum multi-objective design gets a larger S/N ratio of 16.13, while the optimum emission design only gets an S/N ratio of 15.86 S/N. It is in line with the ANOVA test results which show that the optimum multi-objective design is proven to be better than the optimum emission design.

Keywords

• Metallic Catalytic Converter
• Pareto solutions
• Artificial Neural Network
• Genetic Algorithm
• CO Emission
• Engine Performance