基于微分进化算法的双波长光纤F-P传感系统结构优化  

Structural optimization of double wavelength optical fiber F-P sensing system based on differential evolution algorithm

作　　者：单宁[1] 史仪凯[1] 刘霞[1] 

机构地区：[1]西北工业大学机电学院,陕西西安710072

出　　处：《光学精密工程》2009年第2期301-305,共5页Optics and Precision Engineering

基　　金：国家自然科学基金资助项目(No.50275125);航空科学基金资助项目(No.2007ZE53050)

摘　　要：针对光纤法布里-珀罗(F-P)传感器在实际应用中易偏离工作点,造成输出信号衰减,信噪比降低问题,提出了双波长稳定技术,建立了双波长稳定系统的数学优化模型。采用全局搜索能力强,收敛速度快,鲁棒性好,能有效解决复杂优化问题的微分进化算法(DE)进行了全局结构优化。结果表明,DE算法能在较短的运行时间内得到全局最优解,误差<10-3,优化目标下降了89.46%,且运行时间有所减少;证明该算法正确可行,高效可靠,为光纤F-P传感器的结构优化设计提供了实验基础。On the basis of Differential Evolution （DE） algorithm with strong capabilities of global searching, fast convergence speed, good robustness and easy to optimize complex problem,a technology of double wavelength stabilization is put forward to resolve the problems that optical fiber F-P sensor deviates from working point easily and causes the attenuation of output signal and the decrease of Ratio of Signal to Noise（RSN）. The mathematic optimal model of double wavelength stabilization system is established, and structure of the system is optimized by using DE algorithm. The results indicate that the global optimal solution can be obtained in a shorter running time, the object is decreased by 89.46% and calculation error is less than 10.3. The results reported above prove that this algorithm is proper and feasible, and can be used for the structural optimization design of optical fiber F-P sensor.

关 键 词：微分进化算法 光纤传感系统 Fabry-Perot干涉仪 结构优化 全局最优 

分 类 号：TN253[电子电信—物理电子学]