Title :
Nonlinear Modeling of MEMS Deformable Mirror Based on Neural Network
Author :
Li, Jie ; Wu, Peng
Author_Institution :
Dept. of Electr. & Inf. Eng., Wuhan Polytech. Univ., Wuhan, China
Abstract :
The nonlinear response and strong coupling effect of control channel in Deformable Mirrors make it difficult to obtain the desired mirror surface shapes. An efficient nonlinear model of deformation with respect to input voltages is presented using a back propagation neural network (BPNN). The residual relative error of the proposed model shows the improvement of accuracy of an order about 5 as compared to that of linear model, and with no significant increase of time.
Keywords :
backpropagation; micro-optomechanical devices; mirrors; neural nets; optical engineering computing; MEMS deformable mirror; backpropagation neural network; deformation; mirror surface shapes; nonlinear modeling; strong coupling effect; Actuators; Adaptation model; Deformable models; Laser modes; Mathematical model; Mirrors; Shape;
Conference_Titel :
Photonics and Optoelectronics (SOPO), 2011 Symposium on
Conference_Location :
Wuhan
Print_ISBN :
978-1-4244-6555-2
DOI :
10.1109/SOPO.2011.5780582
