基于GA-BP神经网络的铁路技术站作业效率优化研究  

Study on Optimization of Operation Efficiency of Railway Technical Station Based on GA-BP Neural Network

作　　者：孔德扬 王梦杰 高磊 KONG Deyang;WANG Mengjie;GAO Lei(School of Traffic and Transportation,Xi'an Traffic Enginering Institute,Xi'an 710000,China)

机构地区：[1]西安交通工程学院交通运输学院,陕西西安710000

出　　处：《甘肃科学学报》2022年第3期104-111,共8页Journal of Gansu Sciences

摘　　要：为提升技术站作业效率,在充分分析技术站车流组织过程的基础上,提出人为因素是影响技术站作业效率的主要因素。首先选取人员出勤情况、作业车数量、车流时段分布\,班次\,天气5个自变量进行分析并利用单因素分析法对自变量进行量化与筛选,以各单项作业的超时系数构建输出数据评价体系;其次选取GA-BP神经网络作为预测模型以确定最优权值和阈值;最后利用遗传算法对最佳班组人员组合进行计算,并用实例进行验证。结果表明:GA-BP神经网络测试误差值仅为0.17229,可见在有限样本数据的条件下,班组人员组合与作业效率的评价等级有较强烈的非线性关系。研究为提升现场作业效率、合理分配班组人员提供了一定的理论基础。In order to improve the operation efficiency of technical station,on the basis of fully analyzing the organization process of train flow in technical station,it is proposed that human-caused factor is the main factor affecting the operation efficiency of technical station.Firstly,five independent variables,including team composition,train flow and weather,are selected for analysis,and the independent variables are quantified and screened by single factor analysis method.Then,the output data evaluation system is constructed by the overtime coefficient of each single operation.Thirdly,GA-BP neural network is selected as the prediction model to determine the optimal weight and threshold.Finally,the optimal team composition is calculated by genetic algorithm and an example is given to verify it.The results show that the test error of GA-BP neural network is only 0.17229.It can be seen that under the condition of limited sample data,there is a strong nonlinear relationship between the team composition and the evaluation level of operation efficiency,which provides a theoretical basis for improving the operation efficiency and reasonably allocating team members in worksite.

关 键 词：技术站作业效率 人为因素 GA-BP神经网络 遗传算法 最佳班组人员组合 

分 类 号：U29-39[交通运输工程—交通运输规划与管理]