An improved sequential importance sampling method for reliability assessment of renewable power systems with energy storage

Abstract

Due to the temporal coupling constraints of energy storage, the reliability of modern renewable power systems can no longer be assessed by non-sequential methods. However, the existing sequential methods cannot be combined with variance reduction methods as easily as the non-sequential ones. Hence, the power system reliability assessment is faced with incredible computational burden. In this paper, we propose an improved sequential importance sampling method to accelerate the computation of power system reliability indices. The proposed method is composed of two stages, namely the pre-simulation and main simulation stages. In the pre-simulation stage, we design a customized probability distribution parameter optimization process for load curves in the cross entropy based importance sampling method. In the main simulation stage, the derived probability distribution parameters in the pre-simulation stage are converted and further used to carry out a sequential Monte Carlo sampling process to efficiently estimate the reliability indices of renewable power systems with energy storage. The proposed method is validated by numerical tests performed on the modified IEEE-RTS 79 and IEEE-RTS 96 test systems, both of which are integrated with three wind farms, one photovoltaic station and one battery energy storage power station. Results show that the proposed method is much faster than existing sequential importance sampling methods in both test systems because of exploiting the acceleration potential of the load. Therefore, the proposed method can be used to assess the reliability of the renewable power systems with more energy storage devices in the future.