A knowledge-based approach to assembly sequence planning

Assembly planning plays a major role in the manufacturing industry. In order to reduce computational complexity, this paper presents a knowledge-based approach to the assembly sequence planning problem. The CSBAT (connection-semantics-based assembly tree) hierarchy proposed in this paper provides an appropriate way to consider both geometric information and non-geometric knowledge. In this research, the typical or standard CSBAT is applied to a given assembly problem. The structure of the KBASP (knowledge-based assembly sequence planning system) is proposed and there are different ways to construct plans for a CSBAT: by retrieving the typical base, by retrieving the standard base, and by geometric reasoning. The approach proposed in this paper can generate assembly sequences for each CSBAT directly, without the problem of merging plans for different child CSBATs. The application shows that the knowledge-based approach can reduce the computational complexity drastically and obtain more feasible and practical plans.