A Geometric Programming to Importance Sampling for Power System Reliability Evaluation

A novel Geometric Programming (GP) is presented in the first time by the optimization model of importance sampling parameters (ISP) in Variance Minimization (VM) for importance sampling (IS) of power system reliability evaluation. The key point of the proposed method is that the equality constraints of VM optimization model can be relaxed into inequalities because of its special structure, thus a new GP-VM convex optimization model can be built exactly to solve the difficulty of obtaining the optimal ISP. Numerical results of two test systems verify the effectiveness of the proposed method.