If you just use the Desirability Function in JMP as it is, it is not easy to calculate the Pareto Front, which is often referred to as multi-objective optimal design. In other words, it is difficult to find the optimal solution for other weights than the optimal solution for one weight for multiple objective functions. However, many researchers tend to derive many optimal combinations considering various weights and then actively deal with various requirements to select the optimal solution. In order to implement this, pymoo, a multi-objective optimal design module made in Python, was applied, and the JMP approximate model was linked using the COM API provided by JMP to find the optimal combination. After that, the optimal combination was saved in MariaDB, and the MariaDB results were read into JMP to explore various design areas.
