The 1st seminar of the Fall Friday Seminar Series 2022 “Non-iterative Multiple Energy Flow Calculation Using Data-driven Modeling Techniques” by Hang Tian (PhD Student, School of Electrical Engineering, Shandong University, China) was given on October 14th.
The multiple energy flow (MEF) model is inherently nonlinear, which challenges the fast steady-state analysis and the efficient optimization of the integrated energy system (IES). The core of a deterministic multiple energy flow calculation is to solve a set of nonlinear equations. So far, the Newton-Raphson method has been frequently employed to provide an accurate MEF solution. However, in practice, this approach does not guarantee a fully reliable convergence of MEF calculation for a variety of reasons, including 1) the convergence region remains uncertain, and there is still a divergence concern; 2) the convergence performance is initial guess dependent, and inappropriate initial estimations may result in an incorrect solution. The rapidly evolving artificial intelligence technique in the power system/IES demonstrates a promising way to enable non-iterative power flow (PF) and MEF calculation. Linear regression techniques have been applied to extract the latent linear relationship in PF data, formulating linear PF models that prevent convergence concerns while offering a number of computational benefits. Despite the impressive accuracy of the above regressed linear PF models, deep models with multiple hidden layers have demonstrated a greater capacity to extract complicated nonlinear characteristics than linear models. Although data-driven modeling has shown to be effective in PF analysis, its usefulness in IES, which undoubtedly involves more complex features, has yet to be examined. This lecture will present recent research conducted at Shandong University, introducing two new modeling approaches for the non-iterative multiple energy flow calculation. The first one is intended for an integrated mechanism- and data-driven modeling approach for MEF calculation in an electricity-heat system, while the second one offers a data-driven modeling approach for electricity-heat-gas flows based on a stacking strategy integrating multiple heterogeneous learners.
About the author
Hang Tian received the M.S. degree in electrical engineering from University of New South Wales, Sydney, Australia, in 2017. He is currently working toward the Ph.D. degree in electrical engineering at Shandong University, Jinan, China. His research interests include data-driven modelling of multiple energy flows and optimal scheduling of integrated energy system.