Evolutionary optimization of mould temperature control strategies: encoding and solving the multiobjective problem with standard evolution strategy and kit for evolution algorithms

A modern injection mould is a high-technology tool for mass production. The products typically have to meet high quality demands because they are often visible to the user. In order to manufacture high quality parts, the temperature of the mould has to be kept at an appropriate level and the temperature has to be equally distributed within the form. A net of interconnected deephole bores leads a water–oil emulsion to the critical areas for controlling the temperatures in the tool. Usually, the layout of the bores is constructed by means of human expert knowledge. In this paper an objective mathematical measure for the multiobjective quality of a mould temperature control system is introduced. Two different encoding approaches for cooling systems are compared. An approach to solve the complex layout problem using an evolutionary algorithm is described. To show the complexity of this multiobjective problem and to interpret and compare the quality of the existing algorithms, a new software environment called KEA (kit for evolutionary algorithms) is used.