基于光纤陀螺仪的轨道不平顺精度检测及预测  

Precision Detection and Prediction on Orbit Irregularity Based on Fiber Optic Gyroscope

作　　者：高鸽子[1] 孔军 GAO Gezi;KONG Jun(College of Civil Engineering,Xi’an Vocational and Technical College of Railway,Xi’an 710026,China;AVIC Automation Institute,Northwestern Polytechnic University,Xi’an 710600,China)

机构地区：[1]西安铁路职业技术学院,土木工程学院,陕西西安710026 [2]西北工业大学,中航工业自控所,陕西西安710600

出　　处：《微型电脑应用》2023年第10期27-29,38,共4页Microcomputer Applications

基　　金：陕西省教育厅专项科学研究计划项目(21JK0816)。

摘　　要：为了提高轨道不平顺检测精度,通过光纤陀螺仪开展轨道不平顺测试实验,通过支持向量机开展预测分析研究。采用人工拉弦对检测场轨道的左右轨轨向进行测量,并分析测量工作的精度。通过实验数据分析,验证了基于光纤陀螺仪轨检车测量轨向的内、外符合测量精度;轨检小车凭借惯导技术的连续性和立体交汇测量技术的快速性,在检测过程无须停车,可以实现5 km/h的绝对测量和10 km/h的相对测量,能较好地满足目前高铁运输的需求,为轨道的调整养护提供精准的数据指导。建立了支持向量机预测模型,获得了较低均方误差,并且获得了接近1的决定系数,表明采用此方法构建的预测模型满足可靠性要求。In order to improve the accuracy of track irregularity detection,the test experiment of track irregularity is carried out by the fiber optic gyroscope.The predictive analysis research is carried out by support vector machine.The left and right track direction of the track in the detection field is measured by the artificial string pulling,and the accuracy of measurement is analyzed.Through the analysis of experimental data,the accuracy of the inner and outer alignment of a certain fiber optic gyro instrument track inspection vehicle is verified.By virtue of the continuity of inertial navigation technology and the rapidness of three-dimensional intersection measurement technology,the track inspection car does not need to stop in the detection process,and can realize the absolute measurement of 5 km/h and the relative measurement of 10 km/h,which can better meet the current demand of high-speed rail transportation and provide accurate data guidance for the adjustment and maintenance of track.A support vector machine prediction model is established,the low mean square error is obtained,and the determination coefficient is about 1,which indicates that the prediction model constructed by this method meets the reliability requirements.

关 键 词：光纤陀螺仪 轨检小车 轨道平顺性 轨距 支持向量机 

分 类 号：TP38[自动化与计算机技术—计算机系统结构]