An investigation into uncertain mixed energy dispatching strategies for wind power incorporating improved error estimation bandwidth
Feiran Yang Jian Feng Xu Hu College of Information Science and Engineering,Northeastern University,Shenyang,China
DOI: https://doi.org/10.1080/15435075.2024.2400554
IF: 3.206
2024-09-22
International Journal of Green Energy
Abstract:The inherent unpredictability and volatility of renewable energy generation contribute to insufficient prediction accuracy, resulting in substantial operational discrepancies. Consequently, frequent dispatching adjustments of conventional generator sets are necessary. The imperative of mitigating the frequent activation of generator sets and curbing the associated costs stands as a paramount challenge in the seamless integration of wind energy within the extant power system. To address this issue, the present study incorporates the uncertainty associated with renewable energy into the optimization of hybrid power dispatch, aiming to reconcile system reliability with economic optimization. This investigation introduces the wind power prediction error histogram, employing discrete coordinate data to fit a curve. The curve serves to enhance the calculation strategy of optimal bandwidth in non-parametric Kernel Density Estimation. To further characterize uncertainty, a random sampling method is employed to convert uncertain wind power variables into deterministic values. Additionally, the scenario reduction method is utilized to mitigate computational demands. Subsequently, based on quantified uncertainty information, a hybrid variable spinning reserve capacity power dispatching model strategy is formulated and simulated using the CPLEX solver. Experimental results demonstrate that the proposed method diminishes load shedding and enhances power system stability.
energy & fuels,thermodynamics,green & sustainable science & technology