Title：Model Identification for Applications to Power Engineering
Lecturer：Prof. Uwe Kruger (Rensselaer Polytechnic Institute, U.S.)
Time：9:00-11:00 a.m. Jan.6
Place：Room205 in East 3 Building
Empirical models play a key role in the prediction and monitoring of power systems. This presentation outlines various model structures with focus on applications in power grids and wind forecasts to estimate the yield for windfarms. Using measurements for voltage and phase angle, the presentation first reviews some autoregressive and moving average model structure that are stationary and then introduces integrated structures to deal with nonstationary trends, which can be observed particularly in voltage measurements. This is followed by showing how such models can be identified and how to construct random variables that are not serially correlated and can be utilized for monitoring such power grids. The second part of the presentation summarizes the analysis of recorded meteorological data to predict wind speed using ambient temperature, relative humidity, dew point, mean sea level pressure, global horizontal radiation among others. For this, the performance of various model structures and modeling methods are examined and contrasted.
Introduction about the lecturerr：Dr. Uwe Kruger is the professor of Practice, Department of Biomedical Engineering, Rensselaer Polytechnic Institute, Troy NY, U.S.. His researches include developing methods for applications of probability and statistics engineering mathematics, signal processing, data chemometrics, system identification and modeling. He has published over 50 journal articles，5 book chapters and 1 book (published by John Wiley & Sons), which totally have attracted more than 1600 citations, (Scopus, Web of Science, Google Scholar), an h-index of 22. He also has collaborated with some famous industrial companies and produced 2 patents attracting 26 citations in Google Scholar.