Researcher biography

I received B.Eng and M.Eng degrees from Xi'an Jiaotong University (China), M.Eng (research) from Nanyang Technological University (Singapore), and PhD degree from the University of Adelaide (Australia). I have been working in the School of IT & Electrical Engineering, the University of Queensland, Australia since May 2008. Prior to joining the University of Queensland, I have seven years industry experience. From 1997 to 1999, I worked as a firmware engineer in CET Technologies Pte. Ltd., Singapore. I was a development engineer in Singapore Institute of Manufacturing Technology from 1999 to 2003.

I'm a senior member of IEEE and the seceratary of IEEE PES/DEIS Queensland section. I'm also a member of CIGRE Australian Panel D1 and closely working with CIGRE Australian A2 and B1. My research interests include industrial informatics, condition monitoring and diagnosis, high voltage engineering and electrical insulation, power systems, wireless sensor networks, and sensor signal processing. My research work is closely associated with the Australian electricity supply industry.

My current research is "Power System Asset Management" with the focus on (1) sensing and signal processing to improve the visibility of electricity network asset condition; and (2) data mining with uncertain reasoning for various applications of electricity networks with high penetration of enewables. Some of my recent works are as below.

  1. Short-term photovoltaic generation forecasting based on Bayesian network with spatial-temporal correlation analysis . We developed an inference model built upon Bayesian network for a very short-term PV generation forecast. The model utilizes historic PV generation data and weather data and incorporates spatial similarity and temporal correlation to support PV output forecast.
  2. Photovoltaic nowcasting incorporating sky images. We developed a deep learning-based model, which can fully utilize both sky images and historic PV output data. The model learns features from local spatio-temporal information embedded in sky images, global spatio-temporal correlations embedded in PV output datasets of a number of distributed PV systems and weather characteristics embedded in exogenous dataset. The obtained three types of hidden features are aggregated and applied to predict the PV output.
  3. Forecasting of a single household electrical load for home energy management systems. We have developed a consumption scenario-based probabilistic load forecasting (PLF) algorithm to provide the load forecasting of an individual household.
  4. Optimal power flow considering uncertainty set of wind power. We have proposed a risk-based contingency-constrained optimal power flow model, in which an adjustable uncertainty set of wind power is developed with network contingencies explicitly incorporated. The model is capable of securing the network against both wind power fluctuations and contingencies in a probabilistic manner with the optimal balance between operation cost and risk.
  5. Data-driven power system asset remaining useful life estimation. Utilizing various condition monitoring data of power system asset, we applied a state-space model method to asset's remaining useful life estimation. To solve the nonlinear and non-Gaussian model, a particle filtering (Sequential Monte Carlo) approach is adopted. The posterior probability density function of the state variable obtained from the particle filtering is used to determine the asset remaining useful life.

Prospective PhD students are welcome to contact me ( provided you have strong academic records and relevant background and interest in data analytics and power system.