基于观测路径相似性粒子滤波在陀螺数据处理中的研究  

Particle Filter based on Observation Path Similarity is Studied in Data Processing of Gyro

作　　者：贺凯盈 王涛[1] 李燕敏[1] 谢忠俍[1] HE Kaiying;WANG Tao;LI Yanmin;XIE Zhongyu(Shanxi Railway Engineering Vocational and Technical College,Weinan,Shanxi 714099,China)

机构地区：[1]陕西铁路工程职业技术学院

出　　处：《中国锰业》2019年第3期31-36,共6页China Manganese Industry

基　　金：国家自然科学基金项目(41504001);陕西省教育厅科研专项资助(14JK1167);陕西铁路工程职业技术学院科研基金项目(Ky2017-053)

摘　　要：针对磁悬浮陀螺全站仪在极端工作条件下采集的数据易受干扰的问题,在顾及磁悬浮陀螺全站仪的转子电流数据特性基础上,对标准粒子滤波算法进行改进,提出了基于观测路径相似性的粒子滤波算法,3种代表性工程环境下的工程实测数据滤波后,均值和方差等指标均表明该算法能够更有效改善原始观测数据质量,滤波效果优于标准粒子滤波和自适应卡尔曼滤波,能够有效提高陀螺定向的精度和稳定性。In view of the problem,the data collected by the magnetic suspension gyroscope total station under extreme working conditions is susceptible to interference.On the rotor current data characteristics of the magnetic suspension gyroscope total station,the improvement of the standard particle filter algorithm is proposed on the observation path similarity.The particle filter algorithm,the mean and variance of the measured data of the engineering measured data in three representative engineering environments all indicate that the algorithm can improve the quality of the original observation data more effectively.The filtering effect is better than the standard particle filter and the adaptive Kalman filter.It can effectively improve the accuracy and stability of the gyro orientation.

关 键 词：磁悬浮陀螺数据 观测路径相似性 粒子滤波 数据处理 

分 类 号：TB22[天文地球—大地测量学与测量工程]