Course Project Work

The optimum solution gives values of design variables to reach the desired value of the response variable by taking into account the necessary constraints. In real-life problems, precise constraints formulation is very challenging, and implicit constraints are difficult to quantify. The optimum solution will be only as good as the formulation of a problem. When constraints cannot be formulated with confidence, it is necessary to achieve a target value of the response variable, instead of an optimum value. In some cases, for process or performance improvement purposes, new targets are set for a response variable and corresponding values of design variables need to be evaluated. Evaluating the values of the design variables becomes difficult with an increased number of design variables. More difficulty arises if the relationship between the design variables and the response variable is nonlinear. Infinite solutions are possible to reach the target value of the response variable with n number of design random variables. This article discusses a statistical approach (First order reliability method 2) to find suitable values of design random variables to reach a target value of response variable irrespective of the number of design random variables and type of functional relationship (linear or non-linear).