Robust parameter design based on response surface model considering measurement errors

Response surface is an eective method for robust parameter design. Previous response surface methodologies assume that independent variables are measured without errors. However, this assumption might be violated due to the low capability of the measurement system. The present study employs the response surface method for robust parameter design with measurement errors in variables. In addition, an unbiased estimator is presented in the presence of some measurement errors and an optimal setting which is determined to minimize the expected quadratic loss. An example is then illustrated to verify the eectiveness of the proposed approach. The results showed that the proposed method could achieve better operating conditions, despite measurement errors, than the conventional method.