针对传统神经网络构建的光纤通信网络故障数据识别器,在训练进程中易出现局部突出极小值,导致故障数据识别不准确的问题。对光纤的内部结构和通信故障数据的自动识别问题进行研究,提出一种基于免疫识别的神经网络光纤通信网络故障自动识别模型,分析传统的神经网络在识别进程中的弊端,将改进的神经网络结构同免疫识别理论相结合重新构建神经网络自动识别器,并应用于光纤通信网络中,用给出的相应训练方法进行训练,把光纤通信网络的故障数据存储在识别器中,获取故障模式特征,依据识别器的激活状态自动识别光纤通信网络的故障点。实验结果表明,本文提出的基于免疫识别的神经网络光纤通信网络故障检测方法,可准确识别出光纤通信网络故障数据,其识别率明显高于传统的识别方法。 For the fault identification of optical fiber communication network based on the traditional neural network, local prominently minima appear easily in the training process, leading to inaccurate identification of the fault data. This paper studies the automatic identification of the internal structure of optical fiber and communication fault data and proposes an automatic fault identification model of optical fiber communication network based on immune recognition. The traditional neural network is analyzed in the recognition process. The improved neural network The structure is combined with the immune recognition theory to reconstruct the neural network automatic recognizer, which is used in the optical fiber communication network and trained with the given training methods. The fault data of the optical fiber communication network is stored in the recognizer to obtain the fault pattern features , Based on the recognition of the activated state of the fiber optic communications network automatically identify the point of failure. Experimental results show that the proposed fault detection method based on immune recognition neural network can accurately identify the fault data of optical fiber communication network, and its recognition rate is obviously higher than the traditional identification methods.