基于麻雀搜索算法改进的BP神经网络铁路技术站智能化安全评价方法研究  

Intelligent Railway Technology Station Safety Evaluation Methods Based on BP Neural Network Improved by Sparrow SearchAlgorithm

作　　者：李博宇 刘启钢 孙文桥 叶飞 张岩 彭超 LI Boyu;LIU Qigang;SUN Wenqiao;YE Fei;ZHANG Yan;PENG Chao(Transportation&Economics Research Institute,China Academy of Railway Sciences Corporation Limited,Beijing 100081,China;Dispatching Office,China Railway Beijing Group Co.,Ltd.,Beijing 100860,China)

机构地区：[1]中国铁道科学研究院集团有限公司运输及经济研究所,北京100081 [2]中国铁路北京局集团有限公司调度所,北京100860

出　　处：《铁道货运》2024年第10期49-57,共9页Railway Freight Transport

基　　金：中国铁道科学研究院集团有限公司科研项目(2022YJ297)。

摘　　要：为了提升铁路技术站的安全性,提出一种融合主客观评价方法、麻雀搜索算法与BP神经网络模型的铁路技术站智能安全评价方法,可以有效解决铁路技术站生产作业的安全评价问题。通过熵权法确定属性权重,并建立相应的安全综合评价模型生成BP神经网络的训练和测试样本,实现对铁路技术站安全性的智能化评估。考虑到选取BP神经网络关键参数的偶然性和不确定性,借助麻雀搜索算法对网络关键参数进行全局优化求解,进一步提升其准确性。通过测试分析验证所设计方法的可行性和有效性,为改善及优化铁路技术站的安全管理提供了新的思路。In order to improve the safety of railway technology stations,an intelligent railway technology station safety evaluation method that integrates subjective and objective evaluation methods,sparrow search algorithm,and back propagation(BP)neural network model was proposed,which could effectively solve the safety evaluation problem of production operations at railway technology stations.The entropy weight method was used to determine attribute weights,and a corresponding comprehensive safety evaluation model was established to generate training and testing samples for the BP neural network,realizing intelligent safety evaluation of the railway technology stations.In view of the randomness and uncertainty in selecting key parameters of the BP neural network,the sparrow search algorithm was used to globally optimize and solve the key parameters of the network,further improving its accuracy.The feasibility and effectiveness of the designed method were verified through testing and analysis,providing new ideas for optimizing the safety management of railway technology stations.

关 键 词：铁路智能化 技术站 安全评价方法 BP神经网络 麻雀搜索算法 

分 类 号：U294.1[交通运输工程—交通运输规划与管理] U298.3[交通运输工程—道路与铁道工程]