cv

PhD

• Electrical and Computer Engineering
• GPA : 3.9/4
• Major Courses
  • Stochastic Processes, Mathematical Methods for Electrical Engineering, Computational Learning, Optimization for Machine Learning, Information Theory, Electric Power Distribution Systems, Partial Differential Equations, etc.

Master

• Control Science and Engineering
• GPA: 3.57/4
• Major Courses
  • Mathematical Statistics, Stochastic Process, Linear System Theory, Multi-sensor Information Fusion, System Identification, Advanced Graph Theory, Big Data and Deep Learning, etc.

Bachelor

• Automation
• GPA: 3.5/4
• Major Courses
  • Mathematical Analysis, Linear Algebra and Analytic Geometry, Probability Theory and Mathematical Statistics, Data Structures and Algorithms, Signals and Systems, Operations Research, Digital Signal Processing, Computer Principle and Embeeded System Design, Digital Image and Video Processing, Pattern Recognition, Numerical Analysis and Algorithms, Introduction to Artifical Intelligence, etc.

Exploring the Design Space of Diffusion Bridge Models via Stochasticity Control

• Diffusion bridge models effectively facilitate image-to-image (I2I) translation by connecting two distributions. However, existing methods overlook the impact of noise in sampling SDEs, transition kernel, and the base distribution on sampling efficiency, image quality and diversity. To address this gap, we propose the Stochasticity-controlled Diffusion Bridge (SDB), a novel theoretical framework that extends the design space of diffusion bridges, and provides strategies to mitigate singularities during both training and sampling. By controlling stochasticity in the sampling SDEs, our sampler achieves speeds up to $5 \times$ faster than the baseline, while also producing lower FID scores. After training, SDB sets new benchmarks in image quality and sampling efficiency via managing stochasticity within the transition kernel. Furthermore, introducing stochasticity into the base distribution significantly improves image diversity, as quantified by a newly introduced metric.

Generating Synthetic Load Data with Physics-informed Time-series Diffusion model

• The physical model was decompose to a deterministic process, which is further embedded into conditional diffusion model for mixed load signal generation. The numerical study results demonstrate the superior generative quality.

Data-Driven Control, Optimization in Power Distribution Networks

• We provided a literature survey of recent data driven optimization and decision-making algorithms in power distribution networks. We summarized the related algorithms and divide those algorithms into four categories: mathematical optimization, learning-assisted optimization, physics-informed learning and end-to-end learning.

Learning Power System Dynamics with Neural Ordinary Differential Equations

• We propose a novel framework that employs Neural Ordinary Differential Equations (ODEs) to learn complex power system dynamics from noisy measurements. The numerical study results demonstrated the superior accuracy of the proposed model over the baseline neural network (NN) and its robustness against measurement noise. Furthermore, the analytics results verified the generalization performance across different fault durations and locations.

Physics-informed Learning for Power System Dynamics

• We proposed a Nearly-Hamiltonian neural network to predict transient trajectories and dynamic parameters of the power system by embedding energy conservation laws in the proposed neural network architecture. The numerical study results on the single machine infinite bus system show that the proposed model produces accurate system trajectories and damping coefficient predictions.

• Shaorong Zhang, Yuanbin Cheng, Xianghao Kong, and Greg Ver Steeg, "Exploring the Design Space of Diffusion Bridge Models via Stochasticity Control", https://arxiv.org/abs/2410.21553, under review, 2024.
• Shaorong Zhang, Yuanbin Cheng, Nanpeng Yu, Generating Synthetic Net Load Data with Physics-informed Diffusion Model, https://arxiv.org/abs/2406.01913, under review, 2024.
• Shaorong Zhang, Nanpeng Yu, Patricia Hidalgo-Gonzalez, et al. "Data-Driven Control, Optimization, and Decision-making in Active Distribution Networks," under review, 2024.
• Shaorong Zhang, Koji Yamashita, and Nanpeng Yu, "Learning Power System Dynamics with Neural Ordinary Differential Equations," IEEE PES General Meeting, 2024.
• Shaorong Zhang and Nanpeng Yu, " Learning Power System Dynamics with Nearly-Hamiltonian Neural Networks," IEEE PES General Meeting, 2023.