Objective Function Selection for Array Optimization Using Principal Component Regression

The definition of objective functions is vital when optimizing parameters of an array or an acoustic field. However, the present selections are often qualitative or empirical, this may lead to imprecise optimization. In this paper, an objective function selection framework is proposed based on principal component regression method (PCR). A regression equation is built among inspection results and acoustic field characteristics. The remarkable acoustic characteristics are selected as objective functions in optimization. An inspection example using a linear array is simulated, the effectiveness of the framework is proven by the simulation. This work can be applied to medical ultrasonic and non-destructive testing domains.