BP神经网络用于光镊力的非线性修正  

Application of BP neural network to nonlinearity correction of optical tweezer force

作　　者：王自强[1] 李银妹[1] 楼立人[1] 魏衡华[2] 王忠[1] 

机构地区：[1]中国科学技术大学理学院,安徽合肥230026 [2]中国科学技术大学信息学院,安徽合肥230026

出　　处：《光学精密工程》2008年第1期6-10,共5页Optics and Precision Engineering

基　　金：国家自然科学基金资助项目(No.10474094)

摘　　要：提出了运用BP神经网络法对光镊力的测量和标定进行非线性修正。针对BP算法收敛慢的缺点,采用LM算法改进BP神经网络算法。结合常用的流体力学法进行了实验研究,证实了所提出的方法使光镊力的测量范围扩大了30%。与多项式拟合方法相比,该方法的精度更高。研究结果表明,采用神经网络算法在不需增加或更换实验设备的前提下,可以有效地提高光镊系统的性能指标。The Back-Propagation （BP） neural network was proposed to correct the nonlinearity and to optimize the force measurement and calibration of an optical tweezer system. Considering the low con- vergence rate of the BP algorithm, the I.evenberg Marquardt（LM） algorithm was used to improve the BP network. The proposed method was experimentally studied for force calibration in a typical optical tweezer system using the hydromechanic methods. The result shows that with the nonlinearity correction using BP neural network, the range of force measurement of an optical tweezer system is enlarged by 30 % and the precision is also improved comparing with polynomial fitting method. It is demonstrated that the nonlinearty correction by neural network method improves effectively the performance of optical tweezers without adding or changing measuring system.

关 键 词：光镊 BP神经网络 非线性修正 

分 类 号：TN249[电子电信—物理电子学] TP183[自动化与计算机技术—控制理论与控制工程]