Mean Value Method for Reliability Based Optimization Revisited 2004-01-1126
The use of mean value first order second moment (MVFOSM or Mean Value (MV) method) in reliability based optimization (RBDO) is studied in this paper. Through a simple xN type nonlinear function, it is shown that the error of MV depends on the coefficient of variation (cov) of the design variables. When the cov is small, MV may provide accurate approximation even for highly nonlinear constraints with high probability requirement. A simple numerical example is presented to illustrate the results.