HEURISTICS FOR ROUTE GENERATION IN PIPE ROUTE PLANNING

摘要

The paper presents an approach to generate candidates piping routes by genetic algorithm (GA) to support interactive planning of a piping route. The basic ideas and their extensions are described first. It covers the definition of genes to deal with pipe routes, the concept of spatial potential energy, the method to generate initial individuals for GA optimization, the concept of zone in route generation using GA, evaluation of crossover methods, and definition and application of fitness functions. Then, the paper describes further heuristics, which was implemented as a preprocessor to narrow the search space and to improve the system performance. Using comparison of simulation results, validity of the heuristics is discussed and evaluated.