A computation method for probabilistic optimal power flow (POPF) for power systems with large-scale wind farms based on unscented transformation (UT) technique is proposed. UT technique is employed to transform the POPF problem into the deterministic OPF problem with a few samples to be solved with the interior point method. The randomness of the wind farms’ output power and the uncertainty of the bus load power are considered

the calculation accuracy of the probability characteristic parameters of the variables in the model of POPF using this method under different wind power capacities in power systems is discussed. The numerical results of IEEE 30-bus

57-bus

118-bus standard systems and a real system S-1047 indicate that

compared with Monte Carlo (MC) method

the proposed approach for POPF based on UT technique can remarkably reduce the computing time while keeping a high accuracy. It is easy to implement and facilitates handling of the correlation of the random variables

while providing a powerful tool for the solution and practical application of the POPF problem.