FUZZY-NEURAL RELIABILITY MODELLING FOR 330 MM OPTICAL FLATNESS IN HORIZONTAL MEASUREMENT

摘要

A large scale of optical flatness measurement requires difficult and complex process which needs in order to achieve high precision and accuracy. Traditionally, the measurement relies on Fizeau interferometer system horizontally. However, such process is very expensive together with high attempt. This paper proposes a novel scientific method for measuring flatness by calibration based on the experiment database library data. Fuzzy logic and neural network method are developed for the application of the flat surface measurement process. The result is compared with the traditional method of the three-flat test Fizeau interferometer system. One advantage of this method is an ability to predict measurement result in the future.