Date of Graduation
Assembly sequence is a determining factor in the successful design of an assembly system. An optimum sequence can improve the product quality, increase reliability of assembly and reduce assembly costs. This dissertation presents a general method for evaluation of optimum mechanical assembly sequences for a given product. The program is a task-level program, which is written in high-level language C in a knowledge-based system environment. The method uses a relational model of the assembly. The representations of subassemblies are presented by cutting the connection graph model based on geometric constraints. The problem is formulated by using the best-first search method to search a graph or obtain the optimum sequence. The method is a flexible method, which allows the user to pick a criteria according to application conditions. The presented program picks the robot operation time, wasted operation time, reorientation time and fixture changing time as factors based on geometric and reliability constraints. Reliability constraints are determined based on two uncertainties: robot repeatabilities and tolerances of assembly mates. When compared to computer results for models of assemblies in the literature, good agreement was obtained with the results using the method developed in this dissertation for the same models.
Jiang, Jie, "Optimization of robotic mechanical assembly sequence utilizing probabilistic constraints." (1994). Graduate Theses, Dissertations, and Problem Reports. 9111.