DIJU LIU / 刘柢炬

DIJU LIU刘柢炬

Assistant ProfessorSchool of Automation Science and Engineering
South China University of Technology

Trustworthy Embodied Decision Learning

I develop decision-learning methods that connect perception, knowledge, action, and feedback in real physical systems. My current work emphasizes safe offline reinforcement learning, multi-agent coordination, process systems, and embodied agents.

Email Google Scholar ORCID GitHub
Portrait of DIJU LIU
Guangzhou, China

Latest News

  1. Joined South China University of Technology as an Assistant Professor.

  2. EKG-AC was published in IEEE Transactions on Cybernetics.

  3. STI-MCO was published in IEEE Transactions on Neural Networks and Learning Systems.

Trustworthy
Embodied
Decision Learning

My research asks how embodied agents can perceive complex environments, learn with knowledge and feedback, act safely under human oversight, and generalize across physical scenarios.

Application systemsProcess Systems  /  Robotic Systems

01Multisource Context Perception & State Representation
02Knowledge-Enhanced Safe Decision Learning & Optimization
03Embodied Agent Behavior Generation & Regulation for Human-AI Co-management
04Feedback-Driven Closed-Loop Learning & Cross-Scenario Generalization

Selected Publications

View complete record on Google Scholar →
2025
IEEE TCYB

EKG-AC: A New Paradigm for Process Industrial Optimization Based on Offline Reinforcement Learning With Expert Knowledge Guidance

DIJU LIU, Yalin Wang, Chenliang Liu, Biao Luo, Biao Huang

DOI ↗
2025
IEEE TNNLS

Spatiotemporal Topology-Informed Multiagent Reinforcement Learning Framework for Structured Multiprocess Collaborative Optimization

DIJU LIU, Yalin Wang, Chenliang Liu, Biao Luo, Biao Huang

DOI ↗
2024
IEEE TKDE

Scope-Free Global Multi-Condition-Aware Industrial Missing Data Imputation Framework via Diffusion Transformer

DIJU LIU, Yalin Wang, Chenliang Liu, Xiaofeng Yuan, Kai Wang, Chunhua Yang

DOI ↗
2023
IEEE TII

Data Mode Related Interpretable Transformer Network for Predictive Modeling and Key Sample Analysis in Industrial Processes

DIJU LIU, Yalin Wang, Chenliang Liu, Xiaofeng Yuan, Chunhua Yang, Weihua Gui

DOI ↗
2024
IEEE TASE

Blackout Missing Data Recovery in Industrial Time Series Based on Masked-Former Hierarchical Imputation Framework

DIJU LIU, Yalin Wang, Chenliang Liu, Kai Wang, Xiaofeng Yuan, Chunhua Yang

DOI ↗
2024
IEEE TIM

Multirate-Former: An Efficient Transformer-Based Hierarchical Network for Multistep Prediction of Multirate Industrial Processes

DIJU LIU, Yalin Wang, Chenliang Liu, Xiaofeng Yuan, Chunhua Yang

DOI ↗

Experience & Education

South China University of Technology

Assistant Professor, School of Automation Science and Engineering

Central South University

Ph.D. in Control Science and Engineering

University of Alberta

Visiting Ph.D. Student in Process Control

Central South University

B.Eng. in Automation

Selected Honors

  • First Prize, Nonferrous Metals Industry Science and Technology Award2025 · Rank 9 of 16
  • President's Innovation Scholarship, Central South University2023 & 2024
  • First-Class Academic Scholarship, Central South University2022, 2023 & 2024

Academic Service

  • Invited Session Co-Chair23rd IFAC World Congress · 2026
  • Journal ReviewerIEEE/CAA JAS, IEEE TCYB, IEEE TII, IEEE TNNLS, Measurement Science and Technology, and others
  • IOP Publishing Trusted ReviewerPeer-review recognition