Enhanced cumulant method for probabilistic power flow in systems with wind generation

In this paper, an enhanced cumulant method (E-CM) is proposed for use in systems with embedded wind generation. The conventional cumulant method though fast performs poorly for systems with wind generation which have active points away from the mean. To minimize this error, the method proposed in this work treats the output wind power distribution as a mixed distribution such that the linearization of the load flow equation is done around the mean of each component part. The improvement achievable using this scheme is demonstrated using an IEEE 24 bus RTS which has been modified to include a wind farm. Results obtained which are compared with those from the Monte Carlo simulation (MCS) method affirm to the better performance of the proposed method over the conventional cumulant method. To properly fit the performance of the proposed technique amongst other load flow methods, results have also been compared with those from the 5-Point Estimate Method (5PEM) and the Unscented Transforms (UT) method.