Smart grids (SGs) have been widely recognized as an enabling technology for delivering sustainable energy transitions. SGs have a positive effect on the development of the world economy and society. SG construction plays an important role in responding to global climate change and promoting the sustainable development of the world economy and society. Under such a background, this paper attempts to investigate patent collaborations of the SG field in China. Based on the application data of collaborative patents from State Intellectual Property Office (SIPO) in China, this study employs complex network theory and social network analysis (SNA) method and conducts in-depth research on the patent collaboration network of SG field in China. The trend of patent collaboration was examined, the collaboration network of SG-related patents was investigated, the network characteristics, and the network structure were also explored. The results show that the proportion of enterprises participating in patent collaboration is relatively large for SG field in China, the percentage of collaboration relationships formed by different patent applicants varies greatly, and that the State Grid Corporation of China (SGCC) plays an important role in patent collaboration in SG field currently. It can also be found that patent collaboration is an effective form of cooperative innovation among different entities. Finally, three suggestions are provided in response to the existing problems.

Faced with fierce external competition and rapidly growing demand, China’s rail transit equipment industry can advance by strengthening the collaborative performance among industries, universities, and research institutes (I-U-R). This study aims to construct and visualize the I-U-R co-patent networks of China’s rail transit equipment industry to gain a comprehensive understanding of the patterns and dynamics of collaborative innovation. Based on 2876 cooperative patents from the China National Knowledge Infrastructure database (1986–2018), the study applies statistical and social network analysis methods to systematically analyze the full evolutionary trajectory of I-U-R co-patent networks in China’s rail transit equipment industry, from “quantitative change” (scale expansion) to “qualitative change” (structural transformation). The results show that the connections among network innovation participants have become progressively closer, the network has expanded in scale, and distinct developmental stages have emerged. The network structure has evolved from a single-core to a cross-regional, multi-core configuration, with enterprises consolidating their dominant position, occupying the core of the network. This study provides theoretical support for strengthening China’s rail transit equipment industry’s co-patent networks and contributes to both collaborative innovation network theory and co-patent research methodology.

A hypernetwork framework is developed to study technology-organization-region coupling of China’s artificial intelligence innovation system. Based on 8,222 authorized collaborative invention patents from 2010 to 2024, a collaborative innovation hypernetwork is built, and its evolution path is analyzed from both intra-layer and inter-layer perspectives. The results show that region subnetwork becomes more integrated, while organization subnetwork expands quickly but still remains fragmented. Meanwhile, technology subnetwork broadens a lot, and the cross-layer coupling becomes further concentrated around a small set of key technologies, core organizations, leading regions. This study extends patent-based quantitative analysis by putting cross-layer interaction into a hypernetwork framework.

The artificial intelligence (AI) industry has become an important driver of economic development worldwide. Transnational technological collaborative innovation hypernetworks can help explain how AI-related technologies, inventors, and regions are jointly connected in global innovation activities. Based on the hypernetwork theory, this study moves beyond the node and edge homogeneity assumed in traditional single-layer networks and incorporates heterogeneous entities into a unified analytical framework. Then, a hypernetwork model is constructed to analyze transnational AI technological collaborative innovation. Specifically, technologies, inventors, and regions are modeled as three types of nodes, while both intra-layer relationships and cross-layer relationships are represented within the hypernetwork. The results show that (1) the US, GB, DE, CN, IN, FR, and CA are important nodes in the region subnetwork; (2) Justine Gottschlich, Prasoonkumar Surti and Saras Baghsorkhi are important nodes in the inventor subnetwork; (3) G06N3/08 and G06K9/00 are important nodes in the technology subnetwork; (4) the technology-inventor-region combination of G06K9/00, Fan Zhu, and the US has the strongest radiation and diffusion effect in the hypernetwork. This study provides a theoretical basis and practical guidance for understanding the transnational AI technological collaborative innovation.