论文部分内容阅读
受限于智能化发展水平,用传统方法识别海量光纤断点故障数据不可靠,通讯失败情况时有发生,为改善这一现状,提出基于移动学习的海量光纤断点故障数据识别方法。通过分析移动学习基础理论弊端,给定约束条件,简化模型处理流程,对移动学习识别模型识别效果进行优化。根据光纤网络基本通讯框架,分析海量光纤断点故障数据识别难点,设计针对断点故障数据高频信号和低频信号的去噪函数,增大信号转换断接振幅,从而建立最优移动学习数据识别模型。实验结论证明,所提出的识别方法能够准确识别海量光纤断点故障,有效识别率和模型收敛速度要明显优于传统方法。
Limited to the level of intelligent development, the traditional method to identify the mass of fiber breakpoint fault data is unreliable, communication failures occur from time to time, in order to improve this situation, based on the mobile learning mass fiber breakpoint fault data identification method. By analyzing the drawbacks of the basic theory of mobile learning, given the constraints and simplifying the model processing flow, the recognition effect of mobile learning recognition model is optimized. According to the basic communication frame of optical fiber network, it is difficult to identify the fault data of mass fiber breakpoint. The denoising function of high-frequency signal and low-frequency signal for breakpoint fault data is designed and the amplitude of signal switching disconnection is increased so as to establish the optimal mobile learning data identification model. The experimental results show that the proposed identification method can accurately identify the failure of the mass fiber breakpoint, the effective recognition rate and the model convergence rate are obviously better than the traditional methods.