基于优化神经网络的光纤激光器的最优设计  

Optimal design of fiber laser based on optimization neural network

作　　者：申玉霞[1] 李飞[1] 

机构地区：[1]济源职业技术学院,河南济源459000

出　　处：《激光杂志》2017年第2期73-76,共4页Laser Journal

基　　金：河南省科技攻关项目(162102210366);河南省科技攻关计划项目(162102210366)

摘　　要：针对当前静态神经元控制下的光纤激光器波束集成度不高,约束控制不稳的问题,提出基于优化神经网络的光纤激光器最优设计方法,通过分析光纤激光器的控制参量及基本控制规律,对光纤激光器的滤波特性和传输波束的谱特征进行分析,引入光纤激光器波束控制的约束参量,结合变结构PID神经网络进行光纤激光器的分裂波束定向集成,实现激光器的波束优化控制设计。试验结果表明,采用该方法进行光纤激光器设计的波束指向性较好,能量耗散较低,激光波束控制的品质较高。In view of the current static neurons under the control of fiber laser beam integration is not high, unstable constraint control problem, put forward based on the optimization of neural network optimal design method of fiber laser, through the analysis of the control parameter of fiber laser and the basic control law, the filtering characteristics of fiber laser and the transmission spectrum characteristics of the beam were analyzed, and the introduction of fiber laser beam control constraint parameters, combined with the variable structure PID neural network division of fiber laser beam directional integration, realize the laser beam and to optimize the design of the control. Experimental results show that by using this method to design fiber laser beam directivity is good, low energy dissipation, the quality of the laser beam control is higher.

关 键 词：神经网络 光纤 激光器 波束 能量 

分 类 号：TN248.1[电子电信—物理电子学]