1 Introduction
Conceptual design, including creating functions, decompsing functions and structural design for function and subfunction in accordance with the requirements of product in life cycle, is solution procedure for working principle satisfying functional and structural requirements, and is a systematic design process to conceive the carrier for principle of functional structure. Conceptual design is a divergent thinking and innovative process to find various schemes that may exist, and ultimately determine the best one. It is the important part in the product design and the main means to improve product quality, reduce costs and improve competitiveness of enterprises. Key technology in conceptual design includes information modeling, reasoning technology, program evaluation and program visualization. This article focuses on how to select the optimum scheme from many feasible ones. Document[1] advanced the fuzzy synthetic evaluation to select the best mechanism, but did not take into account the complex relationship among factors. Evaluation indexes system for mechanical motion typically includes technical and economic performance, and many other contents, which not only often contains a number of uncertain knowledge and information, but also gets in touch mutually, influences upon each other among factors of different levels or the same level constituting a large system. How to effectively synthesize the connectivity and constraining that exist in different properties, rationally deal with uncertain knowledge and information, reflect the physical nature of objective, is hot issues decision-makers and experts study, which leads to the results that grey system theory is one of the most reliable means[2]. Grey relational analysis, important approach in grey system theory, is the effective means to analyse system with poor data. Grey relational analysis pictures strength, magnitude, sequence among factors through degree of grey relation. Researching geometric correspondence among factors by mathematical method on the basis of data column of factors is the basic ideas. Single-level grey relational analysis hardly reflects the influential degree and scope that impact indexes in complex grey system exerts. As such, establishes grey relational analysis model with multiobjective and Multilevel, which divids problem to be decided into several subsystem on the basis of attribute of impact factors, and then divides subsystem into several impact factors. Finally, Comprehensively evaluates each program according to the ideology of grey relational analysis resulting in improving information efficiency and evaluation accuracy.