Carlos F.M. Coimbra, Ph.D.
Mechanical and Aerospace Engineering, Jacobs School of Engineering and Center for Renewable Resource Integration/Center for Energy Research
University of California, San Diego
La Jolla, CA 92093-0411
Our primary research goal is to develop the highest-fidelity forecasting engines for variable energy resource integration, focusing mostly on solar and wind generation, but also on integrated renewable forecasts (load+solar; load+wind). Our engines are applicable to load forecasts, and can be used in a variety of other applications where stochastic learning offers a substantial advantage over deterministic methods. Our research interests cover many different areas, including: Atmospheric Radiation and Cloud Physics; Fluid Mechanics, Heat and Mass Transfer; Pattern Recognition; Stochastic Learning Methods; Fractional and Variable Order Methods; Nonlinear Chaos Dynamics; Optimization and Regression Methods (GA, ANN, kNN, SVM, SVR, ARMA/ARIMA, and ensembles of these methods), and Image Processing. We have developed solar forecast engines that span the whole spectrum of temporal horizons and spatial resolutions, from intra-minute to multiple days ahead forecasts, and from single point radiometers to continental regions.