Short-Term Transmission Capacity Prediction of Hybrid Renewable Energy Systems Considering Dynamic Line Rating Based on Data-Driven Model
Yi Su, Mao Tan, Jiashen Teh
Abstract
The output capacity of Hybrid Renewable Energy Systems (HRES) is crucial for dispatching plans and spinning reserve capacity, but it's constrained by renewable energy generator output and transmission tie-line capacity. Considering both dynamic line rating for tie-lines and the entire HRES is beneficial due to their susceptibility to micro weather conditions. Unfortunately, considering them collectively for capacity forecasting involves long-term regular fluctuations and short-term uncertainty changes in weather factors, which reduces prediction accuracy. Thus, a novel data-driven model is introduced in this paper to address the aforementioned issue. Initially, Complete Ensemble Empirical Mode Decomposition with Adaptive Noise (CEEMDAN) is employed to segment the capacity sequence into high-, medium-, and low-frequency components. Subsequently, the high-frequency component, characterized by wind-induced randomness, is predicted using Newton-Raphson-Based Optimizer (NRBO) - Bidirectional Gated Recurrent Unit (BiGRU); the medium-frequency component, reflecting seasonal regularities, is forecasted using Seasonal Autoregressive Integrated Moving Average (SAIMA); and the smooth and periodic low-frequency component is anticipated using Multivariable Linear Regression (MLR). Finally, the predictions from these models are combined to derive the ultimate predictive value. Case studies demonstrate that comprehensive consideration of transmission tie-lines equipped with DLR, as well as HRES, can enhance the external output capability of HRES, especially during periods of abundant wind resources. The proposed data-driven model can capture high-frequency fluctuations, medium-frequency periodicity, and low-frequency trends in capacity to enhance prediction accuracy.