摘要
A configuration space (-space) method was developed in a previous study to support the automation of the layout design of cooling systems. The main function of the -space is to capture all of the feasible geometry of the layout design so that a genetic algorithm (GA) can be used to explore the design space. This paper reports a further extension of the -space method in two respects. The new method overcomes the previous limitation of the non-variability of the design topology. By capturing useful information generated during the computation of the -space, the new method enables the GA to explore both the geometry and the topology of the design during the evolutionary process. A new technique is also developed that enables the method to handle user-defined constraints on the layout design. These constraints are captured by additional -spaces, and a special procedure is developed for computing the -spaces of cooling systems and generating candidate designs so that the resulting design always satisfies the constraints. Vigorous proof of the validity of the new method is provided, and a case study is presented to demonstrate the capability of the method in generating a practical layout design for a cooling system.