Automatic stable assembly sequence generation and evaluation

Stability problems in mechanical product assembly sequence planning have drawn a great amount of research interest in recent years. Most proposed methodologies involve complex geometric and physical analyses. As a result, even for a simple structure, it is difficult to take all the important criteria into account and find efficient solutions. It is also difficult to find an adequate product model for stable assembly analysis. In this paper, three matrices are used to represent the required assembly information for a product model. Methods for using the matrix-based product model and a genetic algorithm, in forward assembly sequence planning, are developed. Stability, contact, and reorientation measures are incorporated into the fitness function of the genetic algorithm based assembly planner. Potential assembly plans are also checked for geometric feasibility. As a result, the planner finds enhanced product assembly sequences, with respect to the initially supplied sequences, that are improved for stability, that require fewer assembly direction reorientations, and that meet product geometric assembly constraints.