Design of Graph-Based Evolutionary Algorithms: A Case Study for Chemical Process Networks Michael Emmerich emmerich@ls11.cs.uni-dortmund.de Center for Applied Systems Analysis, Informatik Centrum Dortmund (ICD/CASA), Joseph von Fraunhofer Straße 20, 44227 Dortmund, Germany Monika Gr¨ otzner groetzner@ltt.rwth-aachen.de Department of Technical Thermodynamics, Technical University Aachen (RWTH), Schinkelstrasse 8, 52062 Aachen, Germany Martin Sch ¨ utz martin.schuetz@nutechsolutions.de Center for Applied Systems Analysis, Informatik Centrum Dortmund (ICD/CASA), Joseph von Fraunhofer Straße 20, 44227 Dortmund, Germany Abstract This paper describes the adaptation of evolutionary algorithms (EAs) to the structural optimization of chemical engineering plants, using rigorous process simulation com- bined with realistic costing procedures to calculate target function values. To represent chemical engineering plants, a network representation with typed ver- tices and variable structure will be introduced. For this representation, we introduce a technique on how to create problem specic search operators and apply them in stochastic optimization procedures. The applicability of the approach is demonstrated by a reference example. The design of the algorithms will be oriented at the systematic framework of metric- based evolutionary algorithms (MBEAs). MBEAs are a special class of evolutionary algorithms, fullling certain guidelines for the design of search operators, whose ben- ets have been proven in theory and practice. MBEAs rely upon a suitable denition of a metric on the search space. The denition of a metric for the graph representation will be one of the main issues discussed in this paper. Although this article deals with the problem domain of chemical plant optimization, the algorithmic design can be easily transferred to similar network optimization prob- lems. A useful distance measure for variable dimensionality search spaces is sug- gested. Keywords Evolutionary algorithms, genetic programming, chemical process optimization, net- work representations, metric-based evolutionary algorithms, minimal moves. 1 Introduction Chemical plants contain multiple chemical engineering devices (e.g., pumps, distilla- tion columns, and chemical reactors) that spread a complex net of connecting material, heat, and information streams. They serve to perform a chemical process during which raw materials are converted into desired products. c 2001 by the Massachusetts Institute of Technology Evolutionary Computation 9(3): 329-354