Probabilistic Power Flow Considering Wind Speed Correlation of Wind Farms

The development of renewable energy tends to introduce uncertainty into power system operation. An obvious example is wind power, which in many systems now plays an important role in the overall generation mix. Since wind speed correlation tends to have a significant impact on the operation of power systems, a method of simulating correlated wind speed in different wind farms for modeling is of importance. This paper proposes using an extended Latin hypercube sampling algorithm to simulate correlated wind speeds for different wind farms when solving probabilistic power flow problems. The proposed method employs rank numbers of the sampling points to generate correlated wind speed samples of different wind farms, which avoids generat ing negative wind speed values in transforming uncorrelate d samples to correlated samples, thus improving sampling accuracy. Simulation results using test systems show that the proposed method is effective, simple and easy to be applied to other correlated renewable energy state sampling problems. The illustrations also show the necessity of taking wi nd speed correlations into consideration in probabilistic power flow.