基于多传感信息融合的轨道线形检测
|
摘要
多传感信息融合是实现轨道线形高精度检测的重要方法,而加速度计和陀螺仪是多传感信息融合中的关键传感器;为了解决加速度计和陀螺仪存在累积误差导致测量精度较低的问题,提出一种基于多传感信息融合的轨道线形检测方法;基于捷联惯性系统和双目视觉的测量原理,建立了双目视觉与惯性测量结合的多传感数据融合模型,并利用扩展卡尔曼滤波实现了双目视觉、加速度计和陀螺仪测量信息的融合,提高轨道线形检测精度;通过实验进行验证,结果表明:基于多传感信息融合方法的测量精度比惯性测量方法提高了近9倍,且测量所得坐标在三个方向上的最大位移绝对误差不超过0.536mm,可有效实现高精度轨道线形检测。
Multi-sensor information fusion is an important method of high-precision track detection,and accelerometer and gyroscope are the key sensors in multi-sensor information fusion.In order to solve the problem of low measurement accuracy caused by the accumulative error of accelerometer and gyroscope,a method of track line detection based on multi-sensor information fusion is proposed.Based on the measurement principle of strapdown inertial system and binocular vision,a multi-sensor data fusion model combining binocular vision,gyroscope and accelerometer is established,and the integration of binocular vision,accelerometer and gyroscope measurement information is realized by using extended Kalman filter to improve the accuracy of track line detection.The experimental results show that the measurement accuracy based on the multi-sensor fusion method is nearly 9 times higher than that of the inertial method,the maximum displacement error of the measured coordinates in the three directions is no more than 0.536 mm,which can realize the high-precision track line shape detection.
Multi-sensor information fusion is an important method of high-precision track detection,and accelerometer and gyroscope are the key sensors in multi-sensor information fusion.In order to solve the problem of low measurement accuracy caused by the accumulative error of accelerometer and gyroscope,a method of track line detection based on multi-sensor information fusion is proposed.Based on the measurement principle of strapdown inertial system and binocular vision,a multi-sensor data fusion model combining binocular vision,gyroscope and accelerometer is established,and the integration of binocular vision,accelerometer and gyroscope measurement information is realized by using extended Kalman filter to improve the accuracy of track line detection.The experimental results show that the measurement accuracy based on the multi-sensor fusion method is nearly 9 times higher than that of the inertial method,the maximum displacement error of the measured coordinates in the three directions is no more than 0.536 mm,which can realize the high-precision track line shape detection.
引文
[1]万传风,魏庆朝.城市轨道交通运行评价指标体[J].中国安全科学学报,2006,16(5):79-83.
[2]王志勇,朱洪涛,胡立峰,等.便携式轨检仪研究[J].铁道建筑,2014,2014(9):117-120.
[3]罗峰.轨道检测方法与线路维修的关系研究[J].居舍,2018(25):209-210.
[4]熊仕勇.轨道不平顺检测系统中关键技术研究[D].成都:西南交通大学,2018.
[5]赵洁.基于多传感器信息融合的轨道缺陷在线检测方法的研究[D].成都:西南交通大学,2015.
[6]赵有明.高速铁路基础设施服役状态检测技术[J].铁道建筑,2015(10):1-6.
[7]吴哲明,孙振国,张文增,等.基于惯性测量单元旋转的陀螺漂移估计和补偿方法[J].清华大学学报(自然科学版),2014(9).
[8]孟环标.双目视觉测量系统误差分析与控制[D].济南:山东大学,2012.
[9]魏世斌,刘伶萍,赵延峰,等.高速轨道检测系统[J].铁路技术创新,2012(1):29-32.
[10]张磊.轨道沉降的连续运动测量模型研究[D].上海:上海工程技术大学,2016.
[11]朱文发,柴晓冬,郑树彬,等.基于捷联惯性系统的轨道长波不平顺检测[J].城市轨道交通研究,2012,15(11):87-90.
[12]郑莹.铁路轨道数码影像的纠正与拼接[D].成都:西南交通大学,2008,1-7.
[13]郑树彬,彭乐乐,李立明,等.双目视觉与惯性信息融合的轨道线形检测[J].振动·测试与诊断,2018,38(2):394-400,425-426.
[14]罗伟,郑树彬.轨道空间线形的运动检测算法[J].上海工程技术大学学报,2014,28(3):202-206.
[15]马颂德,张正友.计算机视觉[M].北京:科学出版社,1997.
[16]陈丽君,柴晓冬,郑树彬,等.基于扩展Kalman滤波的轨道线形检测研究[J].测控技术,2018,37(6):85-89.
[17]Mirzaei F M,Roumeliotis S I.A Kalman filter based Algorithm for IMU-camera Calibration:observability analysis and performance evaluation[J].IEEE Transactions on Robotics,2008,24(5):1143-1156.
[18]Kumar S,Prakash J,Kanagasabapathy P.A critical evaluation and experimental verification of extended Kalman filter,unscented Kalman filter and neural state filter for state estimation of three phase induction motor[J].Applied Soft Computing,2011,11(3):3199-3208.
[19]Martinelli A.Vision and IMU data fusion:closed-form solutions for attitude,speed,absolute scale and bias determination[J].IEEE Transaction on Robotics,2012,28(1):44-60.
[2]王志勇,朱洪涛,胡立峰,等.便携式轨检仪研究[J].铁道建筑,2014,2014(9):117-120.
[3]罗峰.轨道检测方法与线路维修的关系研究[J].居舍,2018(25):209-210.
[4]熊仕勇.轨道不平顺检测系统中关键技术研究[D].成都:西南交通大学,2018.
[5]赵洁.基于多传感器信息融合的轨道缺陷在线检测方法的研究[D].成都:西南交通大学,2015.
[6]赵有明.高速铁路基础设施服役状态检测技术[J].铁道建筑,2015(10):1-6.
[7]吴哲明,孙振国,张文增,等.基于惯性测量单元旋转的陀螺漂移估计和补偿方法[J].清华大学学报(自然科学版),2014(9).
[8]孟环标.双目视觉测量系统误差分析与控制[D].济南:山东大学,2012.
[9]魏世斌,刘伶萍,赵延峰,等.高速轨道检测系统[J].铁路技术创新,2012(1):29-32.
[10]张磊.轨道沉降的连续运动测量模型研究[D].上海:上海工程技术大学,2016.
[11]朱文发,柴晓冬,郑树彬,等.基于捷联惯性系统的轨道长波不平顺检测[J].城市轨道交通研究,2012,15(11):87-90.
[12]郑莹.铁路轨道数码影像的纠正与拼接[D].成都:西南交通大学,2008,1-7.
[13]郑树彬,彭乐乐,李立明,等.双目视觉与惯性信息融合的轨道线形检测[J].振动·测试与诊断,2018,38(2):394-400,425-426.
[14]罗伟,郑树彬.轨道空间线形的运动检测算法[J].上海工程技术大学学报,2014,28(3):202-206.
[15]马颂德,张正友.计算机视觉[M].北京:科学出版社,1997.
[16]陈丽君,柴晓冬,郑树彬,等.基于扩展Kalman滤波的轨道线形检测研究[J].测控技术,2018,37(6):85-89.
[17]Mirzaei F M,Roumeliotis S I.A Kalman filter based Algorithm for IMU-camera Calibration:observability analysis and performance evaluation[J].IEEE Transactions on Robotics,2008,24(5):1143-1156.
[18]Kumar S,Prakash J,Kanagasabapathy P.A critical evaluation and experimental verification of extended Kalman filter,unscented Kalman filter and neural state filter for state estimation of three phase induction motor[J].Applied Soft Computing,2011,11(3):3199-3208.
[19]Martinelli A.Vision and IMU data fusion:closed-form solutions for attitude,speed,absolute scale and bias determination[J].IEEE Transaction on Robotics,2012,28(1):44-60.