Latin Hypercube Sampling with Evolutionary Algorithm for Static Security Risk Assessment

Due to correlation coefficient matrix of initialized samples are not always positive definite, this paper presents the improved Latin Hypercube Sampling (LHS) methods with Evolutionary Algorithm (EA) to control correlation and handle power system probability analysis problem. To deal with the non-positive definite correlation matrix, an improved median Latin hypercube sampling with evolutionary algorithm (EA) called MLHS-EA into Monte Carlo simulationis proposed and investigatedusing IEEE 118-bus system with wind farms. This paper also discusses the misunderstandings about the non-positive definite correlation matrix and application of LHS in power system probabilistic analysis. With the proposed method in this paper, the correlation can be controlledmore effectively than previous LHS methods. The accuracy of LHSfor the static security assessment can also be improved further for solving the probabilistic analysis problem in power system. The effectiveness of the method is validated with the Matlabsimulation results.