智能制造企业创新效率关乎我国未来制造业的全球地位，对于加快发展现代产业体系，巩固壮大实体经济根基具有重要作用。利用超效率DEA模型和Tobit回归方程，测算智能制造企业两阶段创新效率及其影响因素，结果表明，我国智能制造企业创新效率呈逐年上升趋势，但涨幅较小，还存在很大提升空间；高研发高转化和低研发低转化类智能制造企业数量最多，电气机械和器材制造业、汽车制造业创新效率优势明显，个体和行业差异均较大；科技水平、资产规模、创新基础环境、股权集中度对两阶段创新效率均具有正向影响；政府支持、人才结构、市场结构对技术研发效率具有正向影响，但对经济转化效率存在负向影响。最后，从制定合理有效的政府支持政策、完善科技投入管理制度、规范市场竞争机制、合理调整创新人才结构4个方面提出提高我国智能制造企业创新效率的政策建议。

Along with the deep integration of digital economy and traditional manufacturing characterized by digitization, networking and intelligence, intelligent manufacturing has become the focus of economic and technological competition among countries around the world. Chinese intelligent manufacturing is confronted by problems such as weak industrial infrastructure, an insufficient supply of data elements, and the relatively lagging application of intelligent technology. Innovation is the primary driving force of the development of intelligent manufacturing, while intelligent manufacturing is rooted in enterprise innovation. Since enterprises are the main driving force of intelligent manufacturing, and "Internet plus" and industrial Internet are important driving forces in the field of manufacturing, this paper aims to explore the allocation structure of innovation resources in intelligent manufacturing enterprises, make scientific measurement of the innovation efficiency of intelligent manufacturing enterprises, and provide references for how to improve the innovation capability of intelligent manufacturing enterprises.#br#This study selects the enterprises from China's intelligent manufacturing demonstration pilot project as a sample, and constructs a two-stage super efficiency DEA model to measure technology development efficiency, economic transformation efficiency, and overall innovation efficiency. The intelligent manufacturing enterprises are divided into four categories: high R&D and high transformation, high R&D and low transformation, low R&D and high transformation, and low R&D and low transformation. Then the factors and degrees that affect the two-stage innovation efficiency of intelligent manufacturing enterprises are analyzed by the Tobit regression model. The results indicate that, first, the innovation efficiency of intelligent manufacturing enterprises has been increasing year by year, but there is still a lot of room for improvement. There are also significant individual differences, and the proportion of intelligent manufacturing enterprises with high R&D and high transformation,low R&D and low transformation is the highest. The two-stage innovation efficiency advantages of the electrical machinery and equipment manufacturing industry and the automobile manufacturing industry are clear, while the textile manufacturing industry is at a relatively low level of innovation efficiency. The efficiency in the technology research and development stage is the main factor determining the overall efficiency of innovation. Second, from the analysis of factors affecting the innovation efficiency of intelligent manufacturing enterprises, government support has the greatest impact on technology research and development efficiency. However, ignoring market demand and blindly increasing research and development output could lead to a decrease in the economic transformation of research and development achievements. A high-quality technological environment and infrastructure have a boosting effect on the R&D generation and economic transformation of intelligent manufacturing enterprises. Large enterprises are more likely to have higher two-stage innovation efficiency than small and medium-sized enterprises. The equity concentration ratio has little effect on two-stage innovation efficiency. An increase in the number of R&D personnel will increase R&D output, but the increased R&D output during the technology R&D stage cannot be timely converted into economic value, and instead will reduce the economic conversion rate. Market competition is beneficial for the R&D output of intelligent manufacturing enterprises, but it has had a significant negative impact on economic transformation efficiency, indicating that the market structure of intelligent manufacturing enterprises in China is not yet sound.#br#Four policy implications are put forward to improve the innovation efficiency of China's intelligent manufacturing enterprises. First, the government should pay attention to the links and quality when providing support to intelligent manufacturing enterprises, appropriately reduce financial support for the achievement transformation process, support key links through various means such as policy subsidies and tax incentives. Second, the government should focus on supporting fields related to China's intelligent manufacturing innovation development, improve the quality of scientific and technological services, and promote the application of innovative achievements to improve the management system for science and technology investment. Third, it is essential for the government to improve the market competition mechanism, encourage the enterprises to strengthen their market competitiveness through research and development capabilities and new products, and crack down on malicious competition through methods such as illegal infringement and price wars. Fourth, it is necessary to reasonably adjust the structure of innovative talents according to the matching degree between R&D personnel and overall innovation resources, and avoid causing redundant talent investment.#br#