Designing a Turbine Blade Cooling System Using a Generalised Regression Genetic Algorithm

The design of a turbine blade cooling system is a multi-objective optimisation problem involving constraints and complex Interaction among its design variables. The aim of this paper is to develop a methodology to optimise this design using Evolutionary Computing techniques. This paper presents Generalised Regression Genetic Algorithm (GRGA) and the mathematical model of a real-life turbine blade cooling system. Even in the presence of variable interaction, the methodology identifies a number of good feasible designs from which one could be finally chosen based on designerʹs preferences. The research also demonstrates that GRGA is capable of optimising a real-life design.