A Coating Design Method Based on Artificial Neural Network

A new optical coatings design method based on Artificial Neural Network (ANN) is proposed. In the automation optimization of optical coatings design, we calculate optimal solution of optical coatings design on the basis of boundary conditions. Design results satisfy the design requirements, because use the new evaluation function. In the automation optimization, the ANN can choose the best weight quickly, and avoid local maximum and minimum. Practical examples show that this method is highly and effective and reliable. The quality of ANN is decided by the number of learning sample. We can get a good ANN to calculate the needed optical coatings if the learning sample is enough. Through learning ANN can get the optimization of optical coatings design. Excessive number of learning sample caused slower speed of calculation and increase time of learning, but quality of ANN is not increase. Insufficient number of learning sample cannot get the good learning result.