Gate positioning design of injection mould using bi-objective micro genetic algorithm

Abstract

The use of a micro genetic algorithm (mGA)-based approach to solve a bi-objective optimization of an injection mould design problem is presented. The advantage of the mGA-based approach is that it requires fewer computational resources than a conventional GA because it has a smaller population than a conventional GA. The main drawback of the mGA-based approach is that design diversity is not secured when multi-modal and multi-objective designs are investigated. To implement the mGA-based bi-objective optimization procedure, the present study proposes a memory set, a filtering process, weight control, and reproduction from the memory set in order to explore new optimal solutions, and identify more-evenly distributed Pareto surfaces. A number of mathematical functions and a typical structural optimization problem are tested to verify the proposed strategies. The approach is subsequently applied to the bi-objective injection moulding design problem of minimizing both the maximum injection pressure and maximum pressure difference between the gate positions in the runner system.