Abstract: Selective-disassembly sequence planning (SDSP) plays a significant role in the maintenance planning of an aircraft, and it is also used during the design stage for analysis of the maintainability of the aircraft. To solve the product SDSP problem efficiently, a product disassembly hybrid graph model, which describes the connection, nonconnection and precedence relationships between the product parts, is established based on the characteristics of disassembly. According to an analysis of the target parts that should be disassembed, a disassembly parts set can be gained. Further, an optimization model is provided to optimize the selective-disassembly sequence. And the solution methodology based on genetic algorithm combined with binarytree algorithm is provided. The genetic algorithm (GA) is advanced in its code, pace and robustness to tackle complex problems with a large number of parameters. And the binarytree algorithm is used to help limit the tendency for premature convergence of the genetic algorithm and to help avoid local optima. Finally, an example is analyzed in detail, and the result shows that the model is correct and efficient.

Key words: selective-disassembly sequence planning, disassembly hybrid graph, connection matrix, precedence matrix, binary-tree algorithm, genetic algorithms