基于BA-WNN的滑行道安全风险预警方法

发布时间：2018-03-23 18:37

  本文选题：风险预警　切入点：预警指标　出处：《中国安全科学学报》2017年08期

【摘要】：为更有效地实现具有复杂性、时变性及非线性的机场滑行道安全风险预警,降低事故发生率,针对小波神经网络(WNN)训练过程易陷入局部最优以及训练不稳定等影响预测准确性问题,采用蝙蝠算法(BA)优化WNN,设计和实现基于BA-WNN的滑行道安全风险预警方法,并将其与BP神经网络(BPNN)、WNN、遗传算法优化小波网络(GA-WNN)等3种方法进行有效性对比。结果表明:BA-WNN方法的预警准确率最高(约为84%),在所有工况下误警率都较低。
[Abstract]:In order to realize more effectively the safety risk early warning of airport taxiway with complexity, time-varying and nonlinear, and reduce the incidence of accidents, Aiming at the problem that the training process of wavelet neural network (WNN) is prone to fall into the local optimum and the training is unstable, the bat algorithm is used to optimize the WNNs, and the method of taxiway safety risk warning based on BA-WNN is designed and implemented. It is compared with BP neural network BPNN and genetic algorithm optimization wavelet network GA-WNN. The results show that the WA-WNN method has the highest early warning accuracy (about 84%), and the false alarm rate is low under all working conditions.
【作者单位】： 武汉理工大学计算机科学与技术学院;武汉理工大学交通物联网技术湖北省重点实验室;武汉理工大学管理学院;
【基金】：国家自然科学基金资助(71271163)
【分类号】：V328;V351.11
，

本文编号：1654689