Changhong Zhao 赵常宏, Associate Professor

Department of Information Engineering

The Chinese University of Hong Kong (CUHK)

Google Scholar

Current team

- Yi Huang 黄奕 (PhD '25, Wuhan University), postdoc

- Zhenyi Yuan 袁臻毅 (PhD '25, UC San Diego), postdoc

- Yujin Huang 黄宇瑾 (BE '23, Tsinghua University), PhD student

Previous students and postdocs

- Bohang Fang 方泊航 (PhD '25), Dartmouth (postdoc)

- Heng Liang 梁恒 (PhD '25), Huawei HK Research Center

- Tong Wu 吴桐 (PhD '21; primary supervisor: Angela Zhang), University of Central Florida

- Runjie Zhang 张润杰 (MPhil '25), PhD student in SEEM, CUHK

- Xinyi Chen 陈心宜 (Graduate RA '22-'23), PhD, Southeast University

- Zexin Sun 孙泽辛 (Graduate RA '22), PhD, Boston University

- Chenxu Wang 王晨旭 (Graduate RA '19-'21), MSc, CUHK; PhD student, CityU HK

- Xinran Liu 刘欣然 (Graduate RA '19-'20), MSc, CUHK; China Southern Power Grid

- Jinyan Su 苏谨言 (Undergraduate RA '22), PhD student, Cornell University

- Xiaojie Li 李晓婕 (Undergraduate RA '21), PhD student, NTU Singapore

- Kaiping Qu 瞿凯平 (postdoc '24-'25), Fuzhou University

- Wanjun Huang 黄婉君 (postdoc '21-'22), Beihang University

- Wei Lin 林伟 (postdoc '21-'22), Chongqing University

- Sidun Fang 方斯顿 (postdoc '20-'21), Chongqing University

Research projects

Joint optimization of distribution network topology and nonlinear power flow.

- Consider an optimal topology and power flow (OTPF) problem in power distribution networks, which jointly optimizes the on/off status of power lines and the generations, loads, and power flows, for more reliable and efficient operation.

- Develop OTPF solution algorithms with improved scalability and optimality, considering unbalanced three-phase nonlinear AC power flow models, uncertainties of renewable generations and loads, and stability indexes without closed-form expressions.

Accelerating large-scale optimization of nonlinear power flow.

- Develop a hierarchical, distributed, spatially recursive algorithm with improved gradient calculation to solve large-scale nonlinear optimal power flow (OPF) problems fast and accurately.

- Integrate hierarchical distributed computing with decentralized neural networks learned from data to predict iterative descent directions of decision variables, for faster solution.

Machine learning for reliable and efficient power system operation.

- Develop a neural-network-based machine-learning method to design decentralized nonlinear frequency and voltage feedback controllers for power systems.

- Steer power system dynamics to a new equilibrium after the operating condition changes, while guaranteeing asymptotic stability and transient safety of frequency and voltage.

- Reduce the maximum frequency and voltage deviations, their variances under noisy power injections and measurements, and control efforts integrated over time.