通过复用已有数字模型，从而提高装配建模的效率，已经成为学术界研究的热点，但CAD领域的装配建模技术依然存在着亟需解决的技术瓶颈：首先是大型装配体设计速度慢、数据量大的问题；其次是基于复用的装配设计过程尚不够智能。针对这些瓶颈问题，本项目拟结合三维零部件检索技术，研究基于复用的快速轻量化装配建模技术，进而解决装配建模过程中大数量所导致的复杂性问题，提高其智能化水平。在已有研究成果的基础上，本项目拟从如下三个方面展开新的研究工作：第一方面是研究基于零件组合的装配检索技术；第二方面是研究基于装配关系的部件推荐技术；第三方面是研究基于标准构件的装配修改技术。本项目的研究成果将被应用在PDM系统中，使得通过检索和推荐已有轻量化模型，实现轻量化装配体的快速构建和便捷编辑，驱动完整装配体的后台自动生成，从而满足装配建模的复杂性和智能化需求，提高新产品的设计效率。

How to improve the efficiency of assembly modeling based on the reuse of existing digital models has been intensively studied in recent years. However, some technical bottlenecks for fast assembly modeling still remain. First, the solutions are not fast enough when dealing with big assemblies with huge data. Second, the algorithms are not smart enough when designing new assemblies. Our project aims to conquer these technical bottlenecks by using 3D model retrieval algorithms. Based on the latest techniques, we will mainly focus on the following three problems. Firstly, we will study the assembly retrieval technique based on part components. Secondly, we will investigate the parts recommendation technique by using machine learning algorithms and the assembly relationships. Thirdly, the assembly modification techniques based on standard parts will be carefully studied. Our research results could be easily embedded in the PDM systems. We expect to design a system that supports fast and smart design and modification of lightweight assemblies based on the retrieval and recommendation models, and then full size assemblies could be generated automatically following the guidance of lightweight results. The efficiency of designing new products is expected to be greatly improved by using our system due to its ability to handle complex assembly modeling.