Department of Climate and Space Sciences and Engineering in the College of Engineering at the University of Michigan


CLASP Team Proposal Selected for Joint Research in Data Science Funding Award

Posted: July 14, 2017

CLASP Team Proposal Selected for Joint Research in Data Science Funding Award

A research proposal submitted by a team of CLaSP researchers and their collaborators at Shanghai Jiao Tong University in China was recently selected for funding by the University of Michigan and Shanghai Jiao Tong University Collaboration on Data Science.

The beginnings of the collaboration go back to 2005, when the University of Michigan and Shanghai Jiao Tong University formed a joint institute “to manage and direct degree-granting programs offered by both universities to students of both nations.” In 2010, U-M and SJTU created a program to provide seed funding for collaborative research to develop new technologies, which initially focused “on research involving reduction in global carbon emissions and climate impact, as well as development of biomedical technologies to promote human health.” The program was later expanded to include nanotechnologies in all areas.

The CLaSP proposal is titled ‘A weather-process and machine learning combined approach to improve solar forecast for PV power generation’, and CLASP Associate Professor Xianglei Huang leads the team. The Co-Investigators are two assistant research scientists in the Climate & Space department, Dr. Xiuhong Chen, and Dr. Cheng Zhou. The SJTU Principal Investigator is Professor Ruzhu Wang from Green Energy Laboratory in the SJTU School of Mechanical Engineering.

The objective of proposed study is to develop a data-driven scheme for the intra-day forecast of solar irradiance at surface and, consequently, the electricity yield of the photovoltaic (PV) panels. A reliable forecast of electricity yield at this timescale is critical for efficient management of PV penetration in the power grid. The CLaSP team has proposed to use both exogenous and endogenous variables to train two machine-learning algorithms for the solar forecast. The validation will be largely relied on the surface observations and PV system maintained by the SJTU collaborators. 

Congratulations, CLaSP team!