Probabilistic Load Flow Analysis Using Randomized Quasi–Monte Carlo Sampling and Johnson Transformation

Abstract

As the fastest growing type of renewable generation, wind power integration has been widely studied to address both the environmental and the energy concerns. The intermittent and stochastic features of wind power, however, cause remarkable uncertainty in operations, resulting in high complexity in system state analysis. It is desirable to evaluate the system conditions as precisely and efficiently as possible. To handle this problem, this paper proposes a novel probabilistic load flow approach by combining randomized quasi–Monte Carlo (RQMC) sampling with Johnson transformation to achieve satisfying accuracy within a low time consumption. For efficient and sufficient sampling, the low discrepancy sequence is scrambled in a fully random strategy, forming the RQMC sampling approach. Furthermore, considering the distribution characteristics and correlation features of wind energy, the Johnson translation system is introduced. Tests on the wind-integrated IEEE 118-bus system and the French high-voltage transmission network show that the proposed approach is able to achieve satisfactory accuracy and efficiency. Different wind profile models, including Weibull distribution and the historical measurements–based probability density functions, can be precisely handled.