Recently, the 17th International Advanced Robotics and Simulation Technology Competition (RST Competition) “Qunzhi Cup” Cluster Intelligence Algorithm Challenge National Final concluded successfully at the Army Engineering University of PLA. The XI team from QMES, NPU, demonstrated solid technical expertise and innovative thinking, emerging as the winner of the national competition’s first prize and an excellence award in the final after being selected through preliminary rounds, national selection, and intense competition among 350 participating teams. This event, which brought together top universities, provided a platform for young students to enhance their professional abilities through hands-on experience. It also served as an innovative engine for advancing robotics and artificial intelligence technology by integrating industry, academia, and research.

This year, the “Qunzhi Cup” competition focused on the optimisation of heterogeneous computing infrastructures and the innovation of cluster intelligence algorithms. Participating teams were required to develop highly adaptable and stable intelligent algorithm models based on domestic computing platforms. The competition not only tested the participants’ programming skills and mathematical modelling capabilities but also emphasised the ability to understand industry needs—specifically, how to translate algorithmic theory into solutions for real-world scenarios, such as smart ocean monitoring and industrial cluster collaboration.

This competition was guided and led by Associate Professor Cao Jing, the head of the QM012403 class of QMES and the Deputy Chief Engineer of the “Jiutian” Heavy Load UAV Scene Application System at the NPU Cultural Heritage Research Institute, along with Instructor Ge Qu, the counsellor for the QMES 2024 class. The team members included five students from the QM012403 class: Yang Jinhui, Liu Haoyuan, Wu Yutong, Zhou Ziyue, and Li Yirui. Throughout the competition, the team exemplified the NPUer spirit of resilience, tenacity, and determination, overcoming numerous challenges. From reviewing professional literature to engaging in discussions and exchanging ideas, the team embraced bold innovation. In the land-based strike event, the team faced significant technical difficulties, such as large-scale heterogeneous UAV swarm collaborative control and complex environmental obstacle avoidance. The team pushed through these obstacles and creatively proposed a solution that integrated machine learning algorithms with traditional path planning algorithms: using machine learning to simulate flight data and predict the optimal path, while traditional algorithms ensured path safety. The team also designed a distributed target allocation algorithm that dynamically assigns tasks based on UAV characteristics and target parameters and established a real-time communication mechanism for coordinated strikes. In addressing issues like low algorithm runtime efficiency and code instability, the team, under the guidance of Li Fayuan from the NPU Engineering Training Centre and Chen Jin Chao from the School of Computer Science, optimised the algorithm at the foundational logic level. They introduced parallel computing techniques to enhance runtime efficiency and, through debugging and optimising communication protocols, significantly improved the system’s stability. In the final stage of the competition, the team engaged in friendly exchanges with teams from various universities, ultimately winning the Excellence Award.

Through this competition, the team members gained a profound understanding of the power of technology in empowering education and accumulated valuable experience. This outstanding performance is a recognition of the team’s technical capabilities and fully embodies the spirit of the Class Three Materials, which promotes “unity, innovation, dedication to learning, and pursuit of talent”. It also demonstrates the successful integration of NPU’s distinctive “Sanhang” (aeronautics, astronautics, and marine science and technology) with the field of artificial intelligence. This achievement will inspire more students to forge ahead on the paths of professional learning, technological innovation, and practical exploration.

Text: Yang Jinhuan, Cao Jing and Ge Qu

Photo: XI Team

Editor: Zhang Hanen and Chen Xinyan

Translator: Shen Xinyi

Reviewer: Cheng Yin