Resumen
System design in an aircraft is still a costly, manual and iterative approach. One major cost driver of changes in system installation are design efforts for creating new pipes in an earlier stage and the costs accumulated during the in service life. To reduce these costs and the time to market, an automation approach with an integrated design optimization encoded in graph-based design languages and executable in a design compiler is proposed. To generate the pipe work automatically, a set of input data (e.g., start- and end-points of a pipe with tangents and fixing positions) is given by the user. It also contains, among others, the weightings for the optimization criteria (e.g., length of the pipe resp. the weight vs. the number of bends) to influence the evaluation of the generated pipes and thereby the final solution. As an initial step in the automatic pipe generation process, a route through the installation space is searched. Subsequently, the installation space is simplified and a respective minimal distance to each obstacle which a pipe should satisfy is added. Then for each pipe an initial solution is estimated and each pipe is optimized by a simulated annealing algorithm. At last, all given requirements are automatically verified. A carried out investigation indicates a polynomial runtime behaviour of the algorithm. The capabilities of the newly developed automated piping are demonstrated on the pipe work in an Airbus A320 landing gear bay.