An Axiomatic Approach for “Target Cascading” of Parametric Design of Engineering Systems
暂无分享,去创建一个
Abstract Complex engineering system realization involves finding out design specifications that simultaneously achieve performance objectives at different levels. A common practice in industry is to adopt “Target Cascading” to obtain proper settings of the performance objectives, and find out those design specifications, not necessarily optimal, but satisfying all the desirable component-level, subsystem-level and system-level performance objectives. In this paper, an Axiomatic Approach to “Target Cascading” (AATC) is presented to improve the current “Target Cascading” process. AATC uses axioms to guide the decompositions of performance objectives, and an integration of a hybrid meta-modeling tool and direct synthesis method to enhance both robustness and efficiency. The preliminary results of AATC's industrial applications demonstrate its advantage in improving productivity at the early stage parametric design, especially for complex engineering systems.
[1] Peter Norvig,et al. Artificial Intelligence: A Modern Approach , 1995 .
[2] Stephen C.-Y. Lu,et al. A Model Fusion Approach to Support Negotiations during Complex Engineering System Design , 1997 .
[3] S. C.-Y. Lu,et al. Building Layered Models to Support Engineering Decision Making: A Machine Learning Approach , 1991 .