In the process of system modeling, the customer needs were transformed into functional requirements, which were expressed with metamodel. Based on independent axiom and zigzag mapping mode of axiomatic design, the functional requirements were mapped to design parameters, and the design matrix was created, which was then converted into design structure matrix by identifying the relation among functional requirements and the sensitivity of functional requirements to design parameters. Clustering algorithm was utilized to cluster the design parameters and to group the system components into modules in design structure matrix, and the interface among modules could be identified and system robustness interaction matrix was developed. Then the interaction parameters were considered as controllable factors, and experimental design techniques were utilized to analyze the fluency of interaction parameters on the design goal, if any,that mought result in a robust system. The applications of the approach were illustrated with crane example of the design of a crab system, which proves its effectiveness and feasibility.
