CV
Zhixu Duan
Undergraduate Researcher
Summary
Undergraduate at SMEE, UESTC, majoring in Mechanical Engineering (Rank 1/83). Research Assistant at ReliaLab (UESTC) and EPICLab (SJTU). Two-time National Scholarship recipient (Top 1.8%). Passionate about AI4Reliability, zero/few-shot learning, transfer learning, and LLMs. Currently focusing on robust fault diagnosis and industrial code generation.
Education
- B.S. in Mechanical Engineering2027-09University of Electronic Science and Technology of China (UESTC)GPA: Rank 1/83 (Top 1)
- High School2023-09No.1 Middle School of Huairen
Work Experience
- Research Assistant2025-11 - PresentShanghai Jiao Tong UniversityAdvised by Prof. Linfeng Zhang. Focusing on efficient and precision intelligent computing.
- Research Assistant2023-12 - PresentUniversity of Electronic Science and Technology of ChinaAdvised by Prof. Zuoyi Chen and Prof. Hong-Zhong Huang. Researching AI for Reliability and Safety.
Skills
Research Interests
- AI4Reliability
- Zero/Few-shot Learning
- Transfer Learning
- Large Language Models (LLM)
- Industrial Code Generation
Publications
- Pseudo-fault data enhanced relation network for fault detection and localization in train transmission systems2024Engineering Applications of Artificial Intelligence (EAAI, SCI Q1)First author. A relation network using pseudo-fault data improves train transmission fault detection and localization performance.
- Parallel Relation Network for Intelligent Fault Detection and Localization of Train Transmission Systems with Zero-fault Sample2024IEEE PHM 2024First author. Oral presentation. A PRN model combining RSN and KAN is proposed for train transmission fault detection with few samples, achieving over 98% accuracy.
- IndustryCode: A Benchmark for Industry Code Generation2025Submitted to ICML 2026Contributor. A Benchmark for LLM to carry Industry Code Generation.
- Decoupling Intrinsic Fault Features from Domain Variations via Domain-Attribute Fusion for Unseen-Domain Fault Diagnosis2025Submitted to Mechanical Systems and Signal Processing (MSSP, SCI Q1)First author. Under Review. A domain-attribute fusion model decouples intrinsic fault features, improving unseen-domain diagnosis robustness.
- Unified Health Domain Relation Learning for Train Transmission Systems Fault Detection under Complex Operating Conditions2026Structural Health Monitoring (SHM, SCI Q1)Co-first Author (first author of adviser). A unified health-domain relation learning approach enhances fault detection under complex operating conditions.
- Open-Set Fault Diagnosis Using CLIP with Forward-Reverse Reasoning2026Submitted to Computers in Industry (COMPUT IND, SCI Q1)Second author (first author of adviser). Minor Revision. A CLIP-based forward–reverse reasoning model enable for fault diagnosis.
- Collaborative Teacher-Student Learning: Simulated Domain Attacks for Class-Intrinsic Feature Learning in Multi-Domain Generalized Fault Diagnosis2026Submitted to IEEE Transactions on Industrial Informatics (IEEE TII, SCI Q1)First author. Under Review. A collaborative teacher-student learning framework enhances multi-domain generalized fault diagnosis.
- Two Invention Patents2025
Presentations
- Interview: Take Each Step Steadily, and the Distance Will Unfold2024SMEE Idol InterviewUESTCInterviewed by SMEE of UESTC. Available at: https://mp.weixin.qq.com/s/ADhuZslqoPZVoshV1WONIw
- Uni-Lab Developers’ Offline Workshop2024UESTC-IA × DP TechnologyChengduInterviewed by DP Technology regarding the collaboration. Available at: https://uestc-ia.github.io/events/uni-lab-news.html
Teaching
- Vice President & Founding Core Member2023UESTC Interdisciplinary Association (IA)Role: LeadershipInvited Prof. Dezhong Yao and Prof. Pedro Antonio Valdes Sosa as advisors; built and maintained the association's official website; promoted outreach and research in bioinformatics, bioimaging, bioelectronics, and AI for Life Science.
- Student Mentor for SMEE 2025 Freshmen2025UESTCRole: MentorMentored freshmen under the guidance of Counselor Keyu Chen.
Portfolio
- National Innovation and Entrepreneurship Program2024
Languages
- ChineseNative
- EnglishProfessional Working Proficiency
Interests
- Academic PositionsKey reader of Nature, Potential contributor to Science
- Conference ActivitiesTea break self-service acquisition, Multimodal collection