摘要
轨道健康状态评价技术对于保障列车的运行安全与乘客的舒适性有重要意义,为寻求一种新的轨道设备综合评价方法,实现对轨道健康状态的科学评价,提出一种基于改进模糊C均值聚类和序列最小优化算法(SMO)构建轨道健康状态评估分析模型。该模型首先提出包含轨道几何状态和结构状态的综合评价指标体系;其次采用变异系数法计算评价指标的权重系数并代入模糊C均值聚类法,得到各轨道样本的分类结果;在此基础上,再利用序列最小优化算法通过划分数据对轨道健康状态进行评价;最后通过实例分析对该评价模型进行验证并开展研究。研究结果表明,经模型评价的855个轨道单元评价结果中优良比例为94%,预测效果良好,平均误差为5%,进而验证了该模型的指标体系和评价方法的科学性和合理性,并给出了进一步研究优化的方向。本文对各轨道指标统筹综合评价,为地铁轨道工务管理线路质量评价提供一种新思路,使轨道设备管理变得有序可控,减少人力、物力资源的浪费。
The assessment technology of track health is of great significance to ensure operational safety and passenger comfort.To find a new comprehensive evaluation method for track equipment and achieve a scientific assessment of track health,this paper proposed a model based on improved fuzzy C-means clustering and the Sequential Minimal Optimization(SMO)algorithm.The model began by establishing a comprehensive evaluation index system that encompassed both the geometric and structural conditions of the track.Subsequently,the coefficient of variation method was used to calculate the weight coefficients of the evaluation indicators,which were then utilized in the fuzzy C-means clustering method to classify various track samples.On this basis,the SMO algorithm was applied to evaluate track health status through data partitioning.Finally,the evaluation model was validated and studied through case analysis.classified as excellent,indicating robust predictive performance with an average error of just 5%.This outcome affirmed the scientific validity and rationality of the models index system and evaluation method,while also providing guidance for further research and optimization.This study offers a coordinated,comprehensive evaluation of various track indicators,presenting a new approach for quality assessment in subway track engineering management,thereby making track equipment management orderly and controllable,and reducing the waste of human and material resources.
作者
许以凯
杨艺
张明凯
赵才友
万壮
XU Yikai;YANG Yi;ZHANG Mingkai;ZHAO Caiyou;WAN Zhuang(School of Civil Engineering,Southwest Jiaotong University,Chengdu 610031,China;Key Laboratory of High-speed Railway Engineering,Ministry of Education,Southwest Jiaotong University,Chengdu 610031,China;Chengdu Metro Operation Co.,Ltd.,Chengdu 610031,China)
出处
《铁道标准设计》
北大核心
2024年第11期53-59,共7页
Railway Standard Design
基金
国家重点研发计划项目(2022YFB2603404)
国家自然科学基金面上项目(51978585)。
关键词
地铁
轨道
健康状态评价
变异系数法
模糊C均值聚类
SMO算法
subway
track
health assessment
coefficient of variation method
fuzzy C-means clustering
SMO algorithm
