基于改进非等间距灰色模型和PSVM的轨道质量指数预测  

Forecasting of Track Irregularity Based on Improved Non-equal Interval Grey Model and PSVM

作　　者：马子骥[1] 郭帅锋 李元良[1] MA Ziji;GUO Shuaifeng;LI Yuanliang(College of Electrical and Information Engineering,Hunan University,Changsha 410082,China)

机构地区：[1]湖南大学电气与信息工程学院,湖南长沙410082

出　　处：《铁道学报》2018年第6期154-160,共7页Journal of the China Railway Society

基　　金：国家自然科学基金(61771191);中国博士后科学基金(2014M562100);湖南省自然科学基金(2017JJ2052);教育部产学合作协同育人项目(201701056026)

摘　　要：有效预测轨道不平顺的发展趋势对铁路的养护和管理具有重要意义。根据轨道不平顺发展的趋势性和随机性,本文提出一种将非等间距灰色模型与粒子群优化支持向量机结合的预测方法。利用改进的非等间距灰色GM(1,1)模型预测轨道质量指数(TQI)序列在未来一段时间内的变化,再利用粒子群优化的支持向量机(PSVM)模型对灰色预测值进行纠正,得到较准确的TQI序列,构建出轨道不平顺变化趋势预测模型。分别对沪昆线上行两段线路的轨道不平顺进行预测,结果表明该组合模型的预测精度较高,相对误差分别低至1.03%和2.74%。Effective forecasting of track irregularity development trend is of great significance for the maintenance and management of the railway.According to the trend and randomness of the development of the track irregularity,a forecasting method that combines non-equal interval grey model and particle swarm optimization（PSO）support vector machine（SVM）was proposed.Firstly,the improved non-equal interval grey GM（1,1）model was used to predict the changes in the sequence of track irregularity TQI in the next period of time.Then the particle swarm optimization support vector machine（PSVM）model was used to correct the gray prediction value to give a more accurate TQI sequence to construct the track irregularity trend forecasting models.The results of the forecasting of the track irregularity of the two sections of Shanghai-Kunming line show higher precision of the combined model,with the relative errors as low as 1.03% and 2.74%respectively.

关 键 词：轨道不平顺 非等间距 灰色模型 粒子群优化 支持向量机 

分 类 号：U213.2[交通运输工程—道路与铁道工程]